1 00:00:00,080 --> 00:00:02,000 I'm coming to them and saying, learn this 2 00:00:02,000 --> 00:00:04,160 entirely new way to write TypeScript. 3 00:00:04,160 --> 00:00:06,400 Trust me, bro, it'll be worth it. 4 00:00:06,400 --> 00:00:08,560 And it helps to like have some examples I 5 00:00:08,560 --> 00:00:11,360 can point to. And I think I recall one 6 00:00:11,360 --> 00:00:13,120 team member of the one actually with 7 00:00:13,120 --> 00:00:14,960 prior TypeScript experience. 8 00:00:14,960 --> 00:00:16,880 He he trusted me enough to give it a try. 9 00:00:16,880 --> 00:00:19,680 And he said it really clicked for him 10 00:00:19,680 --> 00:00:21,760 when he was writing tests, actually, 11 00:00:21,760 --> 00:00:24,720 because he mocked some layer or 12 00:00:24,720 --> 00:00:27,120 wrote a test layer for for some service. 13 00:00:27,120 --> 00:00:29,440 And he's like, oh, this is what I can do 14 00:00:29,440 --> 00:00:31,760 with this and like isolate just certain 15 00:00:31,760 --> 00:00:32,880 services and tests 16 00:00:32,880 --> 00:00:34,400 just the core business logic. 17 00:00:34,400 --> 00:00:36,240 And it's super helpful there. But 18 00:00:36,240 --> 00:00:37,520 that was the moment that clicked for him. 19 00:00:37,520 --> 00:00:39,200 I think we've all had sort 20 00:00:39,200 --> 00:00:41,547 of that that moment, you know, that that 21 00:00:41,547 --> 00:00:42,320 is just like, oh, this is 22 00:00:42,320 --> 00:00:43,760 what this is capable of. 23 00:00:43,760 --> 00:00:44,880 This needs to be sort of 24 00:00:44,880 --> 00:00:46,160 everywhere on a code base. 25 00:00:49,600 --> 00:00:52,000 Welcome to Cause & Effect, a podcast 26 00:00:52,000 --> 00:00:53,840 about the TypeScript library and 27 00:00:53,840 --> 00:00:56,160 ecosystem called Effect, helping 28 00:00:56,160 --> 00:00:57,120 engineers to build 29 00:00:57,120 --> 00:00:58,800 production-ready software. 30 00:00:58,800 --> 00:01:00,560 I'm your host, Johannes Schickling, and 31 00:01:00,560 --> 00:01:01,440 I've been building with 32 00:01:01,440 --> 00:01:03,120 Effect for over four years. 33 00:01:03,120 --> 00:01:05,360 With this podcast, I want to help others 34 00:01:05,360 --> 00:01:06,640 understand the powers and 35 00:01:06,640 --> 00:01:08,720 benefits of using Effect. 36 00:01:08,720 --> 00:01:10,640 In this episode, I'm talking to Adam 37 00:01:10,640 --> 00:01:14,400 Rankin, CTO of Warp, a YC-backed payroll, 38 00:01:14,400 --> 00:01:17,520 payments, and HR platform for startups. 39 00:01:17,520 --> 00:01:19,600 In this conversation, Adam shares his 40 00:01:19,600 --> 00:01:21,840 path from Rails to TypeScript and how 41 00:01:21,840 --> 00:01:24,480 Warp incrementally adopted Effect across 42 00:01:24,480 --> 00:01:26,880 a single fast-moving monolith. 43 00:01:26,880 --> 00:01:28,880 Let's get into it. 44 00:01:28,880 --> 00:01:30,320 Hey Adam, it's so nice 45 00:01:30,320 --> 00:01:31,520 to have you on the show. 46 00:01:31,520 --> 00:01:32,320 How are you doing? 47 00:01:32,320 --> 00:01:33,760 Doing well, thanks for having me. 48 00:01:33,760 --> 00:01:34,960 Excited to be here. 49 00:01:34,960 --> 00:01:36,800 Yeah, I'm really excited to 50 00:01:36,800 --> 00:01:38,640 learn more about Warp today. 51 00:01:38,640 --> 00:01:40,480 But before we talk about Warp, do you 52 00:01:40,480 --> 00:01:43,040 mind sharing more about who you are and a 53 00:01:43,040 --> 00:01:45,200 little bit of your background? 54 00:01:45,200 --> 00:01:48,800 Sure, so I'm Adam. I'm the CTO of Warp. 55 00:01:48,800 --> 00:01:51,280 I came to Warp maybe three years ago, 56 00:01:51,280 --> 00:01:53,280 started it with the CEO who 57 00:01:53,280 --> 00:01:55,200 I actually met on Twitter. 58 00:01:55,200 --> 00:01:56,400 Before that, I've only 59 00:01:56,400 --> 00:01:58,080 ever worked at startups. 60 00:01:58,080 --> 00:02:00,960 I have experience building web apps of 61 00:02:00,960 --> 00:02:03,680 all sizes using mostly Ruby on Rails. 62 00:02:03,680 --> 00:02:06,560 And I trained as an engineer in school. 63 00:02:06,560 --> 00:02:07,520 That is awesome. 64 00:02:07,520 --> 00:02:09,600 So you mentioned like most of your 65 00:02:09,600 --> 00:02:12,080 background being in Ruby on Rails. 66 00:02:12,080 --> 00:02:14,080 Can you already foreshadow a little bit 67 00:02:14,080 --> 00:02:15,520 of like how you went from 68 00:02:15,520 --> 00:02:17,920 Ruby on Rails to TypeScript? 69 00:02:17,920 --> 00:02:19,680 Did that happen at Warp? 70 00:02:19,680 --> 00:02:20,880 Did that already happen 71 00:02:20,880 --> 00:02:22,560 as sort of a prior role? 72 00:02:22,560 --> 00:02:24,640 Sure, so I actually started 73 00:02:24,640 --> 00:02:26,560 my career in Ruby on Rails. 74 00:02:26,560 --> 00:02:28,400 I worked at a startup while I was in 75 00:02:28,400 --> 00:02:29,120 school who was 76 00:02:29,120 --> 00:02:31,920 building a web app on Rails. 77 00:02:31,920 --> 00:02:33,520 And I actually didn't know any 78 00:02:33,520 --> 00:02:34,880 other way to build a web app. 79 00:02:34,880 --> 00:02:36,640 It's a great experience, Rails. 80 00:02:36,640 --> 00:02:38,720 It's very opinionated. 81 00:02:38,720 --> 00:02:40,560 Ruby is a very sort of 82 00:02:40,560 --> 00:02:41,760 easy language to write. 83 00:02:41,760 --> 00:02:44,080 I like to say if you know English, you 84 00:02:44,080 --> 00:02:45,680 know Ruby because you're just kind of 85 00:02:45,680 --> 00:02:47,360 like guessing the 86 00:02:47,360 --> 00:02:49,360 method names and they exist. 87 00:02:49,360 --> 00:02:51,280 And then I actually sort of took a long 88 00:02:51,280 --> 00:02:53,440 break from it while I was writing some 89 00:02:53,440 --> 00:02:57,360 Python and various backend services for a 90 00:02:57,360 --> 00:02:59,200 bigger company, a banking company. 91 00:02:59,200 --> 00:03:01,440 And then when it came time to start Warp, 92 00:03:01,440 --> 00:03:02,320 we kind of were 93 00:03:02,320 --> 00:03:03,840 choosing our initial stack. 94 00:03:03,840 --> 00:03:05,440 And we were choosing something that could 95 00:03:05,440 --> 00:03:08,160 make us move as fast as possible and ship 96 00:03:08,160 --> 00:03:09,680 early features to customers. 97 00:03:09,680 --> 00:03:11,840 And at the end of the day, I basically 98 00:03:11,840 --> 00:03:13,520 googled, you know, what do 99 00:03:13,520 --> 00:03:15,920 people use in web apps nowadays? 100 00:03:15,920 --> 00:03:17,440 And it's like, okay, 101 00:03:17,440 --> 00:03:19,520 Next.js has some cool logos. 102 00:03:19,520 --> 00:03:20,640 Let's try this. 103 00:03:20,640 --> 00:03:22,160 And that's sort of how we 104 00:03:22,160 --> 00:03:23,440 started with TypeScript. 105 00:03:24,240 --> 00:03:25,920 I actually want to linger a little bit 106 00:03:25,920 --> 00:03:29,120 more on that because I feel like that 107 00:03:29,120 --> 00:03:31,520 there is still some aspects about like 108 00:03:31,520 --> 00:03:33,920 Ruby on Rails is not bad it is like I'm 109 00:03:33,920 --> 00:03:35,840 not sure numbers wise, I wouldn't be 110 00:03:35,840 --> 00:03:38,800 surprised if it has more usage than ever 111 00:03:38,800 --> 00:03:41,920 before, just because like the well, many, 112 00:03:41,920 --> 00:03:43,840 many people are building software in 113 00:03:43,840 --> 00:03:45,520 many, many different ways. 114 00:03:45,520 --> 00:03:48,720 I think like one of the like North stars of 115 00:03:48,720 --> 00:03:50,880 like what developer experience can be 116 00:03:50,880 --> 00:03:53,520 like that I think in many regards, the 117 00:03:53,520 --> 00:03:56,000 TypeScript ecosystem has not been able to 118 00:03:56,000 --> 00:03:58,880 replicate partially due to just the 119 00:03:58,880 --> 00:04:01,040 constraints of the language, but also 120 00:04:01,040 --> 00:04:03,760 like there's some like vibes and some 121 00:04:03,760 --> 00:04:06,000 like maturity and some like elegance of 122 00:04:06,000 --> 00:04:07,840 the ecosystem that I also 123 00:04:07,840 --> 00:04:10,560 think like kind of culturally is 124 00:04:10,560 --> 00:04:12,320 quite different in the JavaScript and 125 00:04:12,320 --> 00:04:14,960 TypeScript ecosystem and I 126 00:04:14,960 --> 00:04:17,040 don't agree with like all the 127 00:04:17,040 --> 00:04:18,760 perspectives that the Ruby 128 00:04:18,760 --> 00:04:21,920 on Rails creator DHH has. 129 00:04:21,920 --> 00:04:23,280 I still think that there's 130 00:04:23,280 --> 00:04:25,040 like very interesting ideas. 131 00:04:25,040 --> 00:04:27,360 So I'm curious whether you're still kind 132 00:04:27,360 --> 00:04:29,600 of like keeping up with the developments 133 00:04:29,600 --> 00:04:31,600 where it's like Ruby on Rails is going 134 00:04:31,600 --> 00:04:33,440 whether you're still can derive like some 135 00:04:33,440 --> 00:04:35,280 inspiration from that ecosystem. 136 00:04:35,280 --> 00:04:37,040 I haven't kept up with the development 137 00:04:37,040 --> 00:04:39,760 lately. It's funny you say that because 138 00:04:39,760 --> 00:04:41,520 Rails is definitely lingering. It's 139 00:04:41,520 --> 00:04:42,800 definitely still strong. 140 00:04:42,800 --> 00:04:45,520 You know, every I feel like once a month, 141 00:04:45,520 --> 00:04:48,240 you find out like a new website or some 142 00:04:48,240 --> 00:04:51,200 new product that has billions of users or 143 00:04:51,200 --> 00:04:53,360 sorry, millions of users and it's like, 144 00:04:53,360 --> 00:04:55,440 wait, they're on Rails. No way. 145 00:04:55,440 --> 00:04:57,200 So many like of these huge products 146 00:04:57,200 --> 00:04:59,120 started on Rails and I think it's because 147 00:04:59,120 --> 00:05:02,240 it has so much out of the box that you 148 00:05:02,240 --> 00:05:04,960 you need to run like like jobs for 149 00:05:04,960 --> 00:05:07,360 example, you know, you pick up a 150 00:05:07,360 --> 00:05:09,440 TypeScript project and you're like, how 151 00:05:09,440 --> 00:05:11,680 do I queue up some background job and 152 00:05:11,680 --> 00:05:15,040 like respond to my users API request and 153 00:05:15,040 --> 00:05:16,240 in Rails, that's kind 154 00:05:16,240 --> 00:05:18,400 of trivial in TypeScript. 155 00:05:18,400 --> 00:05:21,280 There's no opinionated way to do that. 156 00:05:21,280 --> 00:05:22,560 People are doing all kinds of different 157 00:05:22,560 --> 00:05:24,160 things. You have to kind of build your 158 00:05:24,160 --> 00:05:25,920 stack more and think about it. 159 00:05:25,920 --> 00:05:28,400 I think good DX tends to always win. That's 160 00:05:28,400 --> 00:05:31,040 why we see Next JS on Vercel winning. 161 00:05:31,040 --> 00:05:33,360 Rails is super easy and it kind of 162 00:05:33,360 --> 00:05:35,520 abstracts away the complications of doing 163 00:05:35,520 --> 00:05:38,160 a lot of infrastructure heavy activity. 164 00:05:38,160 --> 00:05:39,680 And I think that's something that the 165 00:05:39,680 --> 00:05:41,200 TypeScript ecosystem in 166 00:05:41,200 --> 00:05:43,200 general is, you know, behind on. 167 00:05:43,200 --> 00:05:45,520 Yeah, I definitely agree. I think what 168 00:05:45,520 --> 00:05:47,280 you've mentioned about like the queuing 169 00:05:47,280 --> 00:05:49,760 system or like any kind of common 170 00:05:49,760 --> 00:05:52,880 canonical thing. Yes, it is possible with 171 00:05:52,880 --> 00:05:55,840 TypeScript, but you like open 10 172 00:05:55,840 --> 00:05:58,000 different blog posts that show you how is 173 00:05:58,000 --> 00:06:00,000 like any find 10 different ways. 174 00:06:00,000 --> 00:06:02,400 Whereas like in the Ruby and Rails world, 175 00:06:02,400 --> 00:06:04,240 like you have 10 blog posts and they all 176 00:06:04,240 --> 00:06:06,400 show the same like there's just like in 177 00:06:06,400 --> 00:06:08,480 one world you're kind of like at ease 178 00:06:08,480 --> 00:06:11,122 where you feel like okay like all roads 179 00:06:11,122 --> 00:06:11,200 where you feel like okay like all roads where you feel like okay like all roads 180 00:06:11,200 --> 00:06:13,360 lead to Rome and like in the TypeScript 181 00:06:13,360 --> 00:06:15,120 world is like wait what like no this is 182 00:06:15,120 --> 00:06:16,640 like entirely different 183 00:06:16,640 --> 00:06:18,800 stack like should I use that? 184 00:06:18,800 --> 00:06:20,720 Oh no, people like in Reddit say like 185 00:06:20,720 --> 00:06:22,480 it's dead. Should I use that? And it's 186 00:06:22,480 --> 00:06:25,120 just like it's a very different world. 187 00:06:25,120 --> 00:06:27,920 And I think there's like a lot positive 188 00:06:27,920 --> 00:06:30,240 to be said about about that where like I 189 00:06:30,240 --> 00:06:32,880 think it's a it's a richer place for 190 00:06:32,880 --> 00:06:35,600 different ideas to like be 191 00:06:35,600 --> 00:06:38,160 be explored at the same time. 192 00:06:38,160 --> 00:06:40,480 But there's also the kind of like 193 00:06:40,480 --> 00:06:43,040 confusion that and sometimes even like 194 00:06:43,040 --> 00:06:45,040 like paralysis of choice that 195 00:06:45,040 --> 00:06:47,360 comes alongside. And 196 00:06:47,360 --> 00:06:49,200 you see a similar thing also in 197 00:06:49,200 --> 00:06:52,000 other language ecosystems, Java obviously 198 00:06:52,000 --> 00:06:54,240 being a big one, but I think in Java, you 199 00:06:54,240 --> 00:06:56,400 have something like Spring etc that has 200 00:06:56,400 --> 00:06:58,880 more of like that that rails kind of 201 00:06:58,880 --> 00:07:02,000 coercion in in one direction. 202 00:07:02,000 --> 00:07:05,200 And I think this is also for if we're not 203 00:07:05,200 --> 00:07:07,120 bringing things back to to Effect, what 204 00:07:07,120 --> 00:07:08,560 we want to talk today about. 205 00:07:08,560 --> 00:07:10,880 I think this is also one of the big 206 00:07:10,880 --> 00:07:13,680 pieces of potential of effect. Yes, it 207 00:07:13,680 --> 00:07:16,080 gives you individual building blocks, but 208 00:07:16,080 --> 00:07:19,120 they all fit together in one holistic 209 00:07:19,120 --> 00:07:22,400 cohesive way of building software that I 210 00:07:22,400 --> 00:07:24,480 think can actually bring a lot of the 211 00:07:25,040 --> 00:07:28,000 nice coherent aspects of something like 212 00:07:28,000 --> 00:07:29,760 Ruby and Rails or something of 213 00:07:29,760 --> 00:07:32,400 like Spring Boot to the the TypeScript 214 00:07:32,400 --> 00:07:34,880 world. And so yeah, and I think we're 215 00:07:34,880 --> 00:07:36,880 going to hear quite a lot about that 216 00:07:36,880 --> 00:07:38,560 today is like in the way how you're 217 00:07:38,560 --> 00:07:40,480 building Warp. So thanks a lot for for 218 00:07:40,480 --> 00:07:42,720 sharing your background and like 219 00:07:42,720 --> 00:07:44,400 which twisted paths you've 220 00:07:44,400 --> 00:07:46,400 taken through the world 221 00:07:46,400 --> 00:07:48,160 of software development. 222 00:07:48,160 --> 00:07:50,560 Maybe before we dive into all things 223 00:07:50,560 --> 00:07:52,240 technical, can you share a little bit 224 00:07:52,240 --> 00:07:54,240 more about Warp? So you mentioned you 225 00:07:54,240 --> 00:07:56,560 started it together with a friend who you 226 00:07:56,560 --> 00:07:57,680 met over over the 227 00:07:57,680 --> 00:07:59,520 internet three years ago. 228 00:07:59,520 --> 00:08:02,080 Like how quickly did things develop? Like 229 00:08:02,080 --> 00:08:03,840 I think today it's no longer just the two 230 00:08:03,840 --> 00:08:06,480 of you, but you I believe like you you 231 00:08:06,480 --> 00:08:08,800 raised a fairly big round recently. 232 00:08:08,800 --> 00:08:11,440 Can you share more about the company and 233 00:08:11,440 --> 00:08:13,680 maybe starting with what is Warp? What is 234 00:08:13,680 --> 00:08:16,240 the product and the goal of Warp? 235 00:08:16,240 --> 00:08:18,800 Yeah, definitely. So we have sort of a 236 00:08:18,800 --> 00:08:21,200 funny origin story. I met the 237 00:08:21,200 --> 00:08:24,400 founder on Twitter in sort of this 238 00:08:24,400 --> 00:08:26,960 pseudonymous era where where people were 239 00:08:26,960 --> 00:08:29,760 using like fake Twitters to sort of meet 240 00:08:29,760 --> 00:08:32,320 each other and he had announced a project 241 00:08:32,320 --> 00:08:33,320 that he was working on 242 00:08:33,320 --> 00:08:34,640 and he raised money for 243 00:08:34,640 --> 00:08:36,240 I thought it was cool. It was actually a 244 00:08:36,240 --> 00:08:39,120 consumer social app. So I had joined it 245 00:08:39,120 --> 00:08:41,440 and and we kind of connected on there 246 00:08:41,440 --> 00:08:43,520 And he had posted on there that he was 247 00:08:43,520 --> 00:08:45,360 looking for engineers and I was sort of 248 00:08:45,360 --> 00:08:48,240 looking for my next thing at the time. 249 00:08:48,240 --> 00:08:51,440 So we actually met each other and agreed 250 00:08:51,440 --> 00:08:53,040 to work together before we knew each 251 00:08:53,040 --> 00:08:55,120 other's names. I always like to tell that 252 00:08:55,120 --> 00:08:56,960 story because it's super funny. I was 253 00:08:57,840 --> 00:09:00,160 visiting my parents at the time 254 00:09:00,160 --> 00:09:03,040 And I was like, oh, I have sort of a job 255 00:09:03,040 --> 00:09:05,280 interview and they were like, you know, 256 00:09:05,280 --> 00:09:08,080 with who? And I was I was like, I don't 257 00:09:08,080 --> 00:09:09,200 know his name, so I 258 00:09:09,200 --> 00:09:10,240 can't really tell you that. 259 00:09:11,840 --> 00:09:13,760 And there was a peak of covid as well, 260 00:09:13,760 --> 00:09:16,800 right? Yes, yes, this is like 2021. 261 00:09:16,800 --> 00:09:18,160 at the time for some 262 00:09:18,160 --> 00:09:20,800 personal context, I was like kind of 263 00:09:20,800 --> 00:09:22,320 playing with the idea of starting my own 264 00:09:22,320 --> 00:09:24,960 thing. I had left my previous role. I was 265 00:09:24,960 --> 00:09:27,200 actually living a nomadic life. I was 266 00:09:27,200 --> 00:09:29,360 living in a different city every month. 267 00:09:29,360 --> 00:09:32,640 So I found this social app. I met the 268 00:09:32,640 --> 00:09:35,360 founder of it. We agreed to work together 269 00:09:35,360 --> 00:09:36,800 on this social app. It 270 00:09:36,800 --> 00:09:41,200 was Swift UI, AWS, Firebase. 271 00:09:41,200 --> 00:09:44,720 And when you build a consumer social app, 272 00:09:44,720 --> 00:09:45,840 you need to monetize it 273 00:09:45,840 --> 00:09:47,760 eventually. And there are only two ways 274 00:09:47,760 --> 00:09:49,600 to do that. And it's to get people to pay 275 00:09:49,600 --> 00:09:52,640 or to get people to see ads. 276 00:09:52,640 --> 00:09:55,200 And if you're not growing with some 277 00:09:55,200 --> 00:09:57,840 insane K factor, then you're not going to 278 00:09:57,840 --> 00:10:00,080 make any money on ads. So we ended up sun 279 00:10:00,080 --> 00:10:01,920 setting the project and sort of going 280 00:10:01,920 --> 00:10:04,080 back to the drawing board and saying, you 281 00:10:04,080 --> 00:10:05,600 know, what can we build? 282 00:10:05,600 --> 00:10:08,080 And again, this was mid covid. We were 283 00:10:08,080 --> 00:10:11,440 using a competitor, payroll product at the 284 00:10:11,440 --> 00:10:14,720 time. And our sort of insight was that, 285 00:10:14,720 --> 00:10:15,920 you know, remote work is not going 286 00:10:15,920 --> 00:10:19,120 anywhere. There's a lot of compliance 287 00:10:19,120 --> 00:10:21,520 work in paying remote teams and paying 288 00:10:21,520 --> 00:10:22,480 people across the globe 289 00:10:22,480 --> 00:10:24,160 and across the country. 290 00:10:24,160 --> 00:10:26,880 There are thousands of tax jurisdictions 291 00:10:26,880 --> 00:10:29,120 like what if we could just take all of 292 00:10:29,120 --> 00:10:32,800 this off the plate of founders and build 293 00:10:32,800 --> 00:10:36,960 a payroll payments HR products tailor 294 00:10:36,960 --> 00:10:38,640 made for startups and founders. 295 00:10:38,640 --> 00:10:41,520 So that sort of was our initial insight. 296 00:10:41,520 --> 00:10:45,760 We joined YC and started talking to 297 00:10:45,760 --> 00:10:48,960 startups and other small customers. It's 298 00:10:48,960 --> 00:10:51,040 like, what do you need? Let us handle it. 299 00:10:51,040 --> 00:10:52,480 And that's sort of how we got started. 300 00:10:52,480 --> 00:10:54,640 And our roadmap has sort of been built 301 00:10:54,640 --> 00:10:57,040 from there. We've like built a lot of the 302 00:10:57,040 --> 00:10:58,880 table stakes features you need. And now 303 00:10:58,880 --> 00:11:01,440 it's very product driven by like 304 00:11:01,440 --> 00:11:02,960 what people are requesting. 305 00:11:02,960 --> 00:11:05,200 And then sort of in conjunction with 306 00:11:05,200 --> 00:11:07,440 the whole covid thing, the LLM 307 00:11:07,440 --> 00:11:09,440 revolution has sort of come along. So 308 00:11:09,440 --> 00:11:12,320 we're now exploring ways to be more AI 309 00:11:12,320 --> 00:11:15,120 native and use these things for 310 00:11:15,120 --> 00:11:16,960 automating different processes that that 311 00:11:16,960 --> 00:11:17,600 we don't want to do. 312 00:11:17,600 --> 00:11:19,440 And our customers don't want to do so. 313 00:11:19,440 --> 00:11:20,960 Hopefully that's that's a good background 314 00:11:20,960 --> 00:11:23,280 on kind of how we started and what we do. 315 00:11:23,280 --> 00:11:25,200 Oh, most certainly. And I want to make 316 00:11:25,200 --> 00:11:28,000 sure that we come back to the AI native 317 00:11:28,000 --> 00:11:30,400 automation part as well as that's another 318 00:11:30,400 --> 00:11:31,120 key interest of 319 00:11:31,120 --> 00:11:32,560 mine currently as well. 320 00:11:32,560 --> 00:11:35,280 But I want to linger a little bit more on 321 00:11:35,280 --> 00:11:37,920 sort of like your personal relationship 322 00:11:37,920 --> 00:11:41,200 to Warp as you went from building a 323 00:11:41,200 --> 00:11:44,640 consumer app to now building a software 324 00:11:44,640 --> 00:11:46,400 for other startups, etc. 325 00:11:46,400 --> 00:11:49,760 I guess much more B2B. And I think there 326 00:11:49,760 --> 00:11:52,560 is kind of two perspectives on like how a 327 00:11:52,560 --> 00:11:55,360 startup founder can end up building 328 00:11:55,360 --> 00:11:57,360 something either kind of out 329 00:11:57,360 --> 00:11:59,520 of love or out of frustration. 330 00:11:59,520 --> 00:12:01,680 What I mean with this, you love something 331 00:12:01,680 --> 00:12:03,840 so much. So you want to build something 332 00:12:03,840 --> 00:12:06,080 that's just like fundamentally good and 333 00:12:06,080 --> 00:12:07,360 like makes it even better. 334 00:12:07,360 --> 00:12:09,440 Or you're so frustrated with an 335 00:12:09,440 --> 00:12:11,520 experience that you say, like, OK, this 336 00:12:11,520 --> 00:12:13,520 needs to be solved. And I think you're 337 00:12:13,520 --> 00:12:15,120 falling into the ladder. 338 00:12:15,120 --> 00:12:17,440 But I think what is tricky about that as 339 00:12:17,440 --> 00:12:20,640 a founder myself, like I had plenty of 340 00:12:20,640 --> 00:12:22,400 like very frustrating experiences 341 00:12:22,400 --> 00:12:25,120 throughout my life where I felt, OK, 342 00:12:25,120 --> 00:12:26,480 there can be something better. 343 00:12:26,480 --> 00:12:29,760 But I also then pictured myself in a life 344 00:12:29,760 --> 00:12:31,920 where I'm not like solving this problem 345 00:12:31,920 --> 00:12:35,280 for 10 years and I would not derive a lot 346 00:12:35,280 --> 00:12:37,920 of joy of like getting and becoming an 347 00:12:37,920 --> 00:12:41,680 expert at tax codes in France. 348 00:12:41,680 --> 00:12:43,520 I'm like frustrated enough with like 349 00:12:43,520 --> 00:12:44,880 barely trying to stay 350 00:12:44,880 --> 00:12:46,320 afloat with the German ones. 351 00:12:46,320 --> 00:12:48,400 So I'm just curious like how your own 352 00:12:48,400 --> 00:12:49,680 perspective is sort of 353 00:12:49,680 --> 00:12:51,600 like has evolved on that. 354 00:12:51,600 --> 00:12:54,160 I'm very glad you're all in solving this. 355 00:12:54,160 --> 00:12:55,520 So I have a better 356 00:12:55,520 --> 00:12:57,920 experience not being exposed to that. 357 00:12:57,920 --> 00:13:00,160 But how you're finding your own balance 358 00:13:00,160 --> 00:13:02,960 of like that frustration that you wanted 359 00:13:02,960 --> 00:13:05,600 to address, like you're flying even 360 00:13:05,600 --> 00:13:07,280 closer to that sun of 361 00:13:07,280 --> 00:13:09,760 frustration to like figure it out. 362 00:13:09,760 --> 00:13:10,560 Maybe you can just 363 00:13:10,560 --> 00:13:12,240 reflect a little bit on that. 364 00:13:12,240 --> 00:13:15,680 Yeah, yeah, it's a funny question because 365 00:13:15,680 --> 00:13:18,000 payroll tax codes, it's like the 366 00:13:18,000 --> 00:13:19,680 unsexiest thing you 367 00:13:19,680 --> 00:13:20,640 could probably imagine. 368 00:13:20,640 --> 00:13:22,720 It's like B2B SaaS payroll. 369 00:13:22,720 --> 00:13:24,960 I have like a few answers to that. 370 00:13:24,960 --> 00:13:28,000 I think one is like the fact that not 371 00:13:28,000 --> 00:13:29,520 many people could imagine themselves 372 00:13:29,520 --> 00:13:32,960 doing it is sort of good for us because 373 00:13:32,960 --> 00:13:35,040 we're happy to kind of go in and like get 374 00:13:35,040 --> 00:13:36,320 our hands dirty and let's 375 00:13:36,320 --> 00:13:37,520 figure out what needs to be done. 376 00:13:37,520 --> 00:13:40,400 And I think the actual joy of building a 377 00:13:40,400 --> 00:13:43,600 product comes from seeing a customer use 378 00:13:43,600 --> 00:13:46,320 it the right way, seeing a customer be 379 00:13:46,320 --> 00:13:49,360 satisfied and not have to worry. 380 00:13:49,360 --> 00:13:51,840 One joke we like to make is that we pride 381 00:13:51,840 --> 00:13:55,360 ourselves on least required user minutes. 382 00:13:55,360 --> 00:13:58,320 So we have some founders that have like 383 00:13:58,320 --> 00:14:00,240 not logged into our product in months and 384 00:14:00,240 --> 00:14:01,600 it's just sort of running in the 385 00:14:01,600 --> 00:14:02,560 background for them. 386 00:14:02,560 --> 00:14:05,520 So we like take a lot of joy in people 387 00:14:05,520 --> 00:14:07,440 not actually having to use our product 388 00:14:07,440 --> 00:14:08,560 because we've become 389 00:14:08,560 --> 00:14:10,960 experts in payroll and tax. 390 00:14:10,960 --> 00:14:14,960 And yeah, I definitely know more about 391 00:14:14,960 --> 00:14:16,880 payroll than I ever thought I would. 392 00:14:16,880 --> 00:14:20,960 But I think that building a intuitive UX 393 00:14:20,960 --> 00:14:23,200 for something that's so boring, I think 394 00:14:23,200 --> 00:14:25,200 with sort of these unsexy industries, 395 00:14:25,200 --> 00:14:29,680 people come to expect unsexy and clunky 396 00:14:29,680 --> 00:14:32,240 products and we're trying to change that 397 00:14:32,240 --> 00:14:34,080 and that's really fun to me. 398 00:14:34,080 --> 00:14:36,000 All right. Yeah, that definitely 399 00:14:36,000 --> 00:14:37,920 resonates. And so like if you think about 400 00:14:37,920 --> 00:14:40,960 this as like this delta between like very 401 00:14:40,960 --> 00:14:43,760 frustrating experience without work to 402 00:14:43,760 --> 00:14:46,720 like joy and delight and like blissful 403 00:14:46,720 --> 00:14:48,720 ignorance of like all the stuff that 404 00:14:48,720 --> 00:14:50,640 needs to happen behind the scenes. 405 00:14:50,640 --> 00:14:53,120 Like if you can get it, delta is probably 406 00:14:53,120 --> 00:14:56,480 pretty wide. So thank you for doing that 407 00:14:56,480 --> 00:14:59,280 work that no one else has to do it. 408 00:14:59,280 --> 00:15:01,040 So yeah, I don't want to get much 409 00:15:01,040 --> 00:15:03,520 closer to tax codes, but I want to get a 410 00:15:03,520 --> 00:15:06,720 slight more intuition and context on like 411 00:15:06,720 --> 00:15:09,120 the people of building Warp. 412 00:15:09,120 --> 00:15:11,280 So I've mentioned that you all raised, I 413 00:15:11,280 --> 00:15:13,680 think, a Series A recently and that 414 00:15:13,680 --> 00:15:15,520 allows you to continue to grow. 415 00:15:15,520 --> 00:15:17,680 So how many people are currently in the 416 00:15:17,680 --> 00:15:18,560 company and how many 417 00:15:18,560 --> 00:15:20,640 of those are engineers? 418 00:15:20,640 --> 00:15:23,520 We have 16 people currently in the 419 00:15:23,520 --> 00:15:24,640 company and our 420 00:15:24,640 --> 00:15:27,280 engineering team right now is four. 421 00:15:27,280 --> 00:15:29,760 So we've we've kind of tried to remain 422 00:15:29,760 --> 00:15:31,760 small, even though we just raised a big 423 00:15:31,760 --> 00:15:32,880 round where we're hiring 424 00:15:32,880 --> 00:15:34,880 for all engineering roles. 425 00:15:34,880 --> 00:15:36,480 So if you're listening to this 426 00:15:36,480 --> 00:15:38,480 joinwarp.com/careers 427 00:15:38,480 --> 00:15:41,120 We like to move very fast and remain sort 428 00:15:41,120 --> 00:15:44,320 of nimble and kind of cut out processes. 429 00:15:44,320 --> 00:15:46,560 Yeah, so we've tried to remain small. 430 00:15:46,560 --> 00:15:47,840 We're very conscious of 431 00:15:47,840 --> 00:15:50,640 growing too fast, but also hiring. 432 00:15:50,640 --> 00:15:52,240 That is awesome. And that definitely 433 00:15:52,240 --> 00:15:54,720 resonates. And like for I think a 434 00:15:54,720 --> 00:15:57,840 contemporary example of like who has done 435 00:15:57,840 --> 00:15:59,760 that really well for a long time is 436 00:15:59,760 --> 00:16:02,960 Linear, where like their tax code 437 00:16:02,960 --> 00:16:04,560 equivalent is task management. 438 00:16:04,560 --> 00:16:08,000 And they've built an absolutely amazing, 439 00:16:08,000 --> 00:16:09,520 delightful experience 440 00:16:09,520 --> 00:16:11,040 for project management. 441 00:16:11,040 --> 00:16:13,200 And I think maybe you're thinking of 442 00:16:13,200 --> 00:16:15,680 yourself like being part of the same 443 00:16:15,680 --> 00:16:17,120 category of like next 444 00:16:17,120 --> 00:16:18,720 generational software. 445 00:16:18,720 --> 00:16:20,880 And the parallel I definitely want to 446 00:16:20,880 --> 00:16:23,440 draw here is they've remained super, 447 00:16:23,440 --> 00:16:26,000 super small for a very long time in 448 00:16:26,000 --> 00:16:27,760 regards to engineering team. 449 00:16:27,760 --> 00:16:29,760 And you mentioning that you're still just 450 00:16:29,760 --> 00:16:31,360 four people like I 451 00:16:31,360 --> 00:16:32,800 definitely applaud you for that. 452 00:16:32,800 --> 00:16:36,160 I think this allows you to just have like 453 00:16:36,160 --> 00:16:39,680 a velocity and agility that is hard to 454 00:16:39,680 --> 00:16:42,080 match when when you're growing and 455 00:16:42,080 --> 00:16:44,640 growing in terms of the engineering team. 456 00:16:44,640 --> 00:16:47,120 I think at some point that is inevitable, 457 00:16:47,120 --> 00:16:50,160 but like remaining small and like keeping 458 00:16:50,160 --> 00:16:52,880 the bar high for engineering talent, I 459 00:16:52,880 --> 00:16:55,440 think is definitely the way to go. 460 00:16:55,440 --> 00:16:58,160 So maybe you can share a little bit more 461 00:16:58,160 --> 00:17:00,480 on like the background of the other 462 00:17:00,480 --> 00:17:03,280 engineers, like when we are now drawing a 463 00:17:03,280 --> 00:17:06,560 bridge to Effect, which has not 464 00:17:06,560 --> 00:17:08,560 too many similarities with 465 00:17:08,560 --> 00:17:11,200 Ruby and Rails in that regard. 466 00:17:11,200 --> 00:17:13,360 That is like more coming from like the 467 00:17:13,360 --> 00:17:15,520 functional programming world, even though 468 00:17:15,520 --> 00:17:16,800 that is not like the core 469 00:17:16,800 --> 00:17:18,560 essence of Effect these days. 470 00:17:18,560 --> 00:17:21,440 This is where it had its origin story and 471 00:17:21,440 --> 00:17:24,080 still draws a lot of people to Effect. 472 00:17:24,080 --> 00:17:26,480 And it requires a little bit of like a 473 00:17:26,480 --> 00:17:30,160 entry ticket to learn a few new things. 474 00:17:30,160 --> 00:17:31,920 So I'm curious, like what was sort of 475 00:17:31,920 --> 00:17:34,080 like the catalyst event for 476 00:17:34,080 --> 00:17:36,320 Warp getting on board with Effect? 477 00:17:36,320 --> 00:17:39,760 Yeah, I guess to say a bit about our 478 00:17:39,760 --> 00:17:41,760 early team, I mentioned Rails. 479 00:17:41,760 --> 00:17:43,680 One of our other core engineers came from 480 00:17:43,680 --> 00:17:44,480 a huge product 481 00:17:44,480 --> 00:17:46,880 running on Rails and React. 482 00:17:46,880 --> 00:17:49,760 And our other team members have come from 483 00:17:49,760 --> 00:17:53,600 a big Python world, especially the guy 484 00:17:53,600 --> 00:17:55,760 who's good at AIs, 485 00:17:55,760 --> 00:17:57,120 written a lot of Python. 486 00:17:57,120 --> 00:17:58,640 And then the last member, I guess I can 487 00:17:58,640 --> 00:17:59,920 just enumerate the whole list 488 00:17:59,920 --> 00:18:02,000 since it's since it's small. 489 00:18:02,000 --> 00:18:03,920 He was a new grad, so he was actually 490 00:18:03,920 --> 00:18:06,560 very interested in Rust and systems 491 00:18:06,560 --> 00:18:07,600 programming languages. 492 00:18:07,600 --> 00:18:09,120 So not actually much 493 00:18:09,120 --> 00:18:12,640 TypeScript before we started Warp. 494 00:18:12,640 --> 00:18:14,720 I think that the way, if I could tell the 495 00:18:14,720 --> 00:18:17,920 story of how I first came across Effect 496 00:18:17,920 --> 00:18:20,880 was that I'm in sort of these online 497 00:18:20,880 --> 00:18:24,640 circles that expose me to things way out 498 00:18:24,640 --> 00:18:26,560 of anything I need to 499 00:18:26,560 --> 00:18:29,120 write B2B SaaS payroll software. 500 00:18:29,120 --> 00:18:31,440 And I'm sure you know a lot of these 501 00:18:31,440 --> 00:18:33,120 communities that they're like heavy 502 00:18:33,120 --> 00:18:34,960 systems programmers. 503 00:18:34,960 --> 00:18:38,000 You'll find some very like eccentric 504 00:18:38,000 --> 00:18:39,520 people writing like 505 00:18:39,520 --> 00:18:41,280 the craziest software. 506 00:18:41,280 --> 00:18:42,480 I'm very interested in 507 00:18:42,480 --> 00:18:44,640 like type systems in general. 508 00:18:44,640 --> 00:18:47,680 And I was writing a lot of random side 509 00:18:47,680 --> 00:18:48,560 projects in Haskell. 510 00:18:48,560 --> 00:18:51,280 Yeah, and I always love that experience. 511 00:18:51,280 --> 00:18:54,080 But I just couldn't justify for the 512 00:18:54,080 --> 00:18:55,280 business like starting to 513 00:18:55,280 --> 00:18:57,040 write something in Haskell. 514 00:18:57,040 --> 00:18:59,360 It would be like such a big 515 00:18:59,360 --> 00:19:00,720 learning curve for people. 516 00:19:00,720 --> 00:19:01,680 It would be like, how 517 00:19:01,680 --> 00:19:03,280 do we deploy this thing? 518 00:19:03,280 --> 00:19:05,520 How are we going to build our entire 519 00:19:05,520 --> 00:19:06,800 pipeline around this? 520 00:19:06,800 --> 00:19:08,480 I can't justify that when you're trying 521 00:19:08,480 --> 00:19:10,240 to just ship features. 522 00:19:10,240 --> 00:19:13,040 But anyway, back to how I found Effect. 523 00:19:13,040 --> 00:19:15,840 I think it was in these circles like some 524 00:19:15,840 --> 00:19:18,880 of these guys obsessed with type systems 525 00:19:18,880 --> 00:19:19,840 are like, why would 526 00:19:19,840 --> 00:19:21,280 anyone ever use TypeScript? 527 00:19:21,280 --> 00:19:22,160 It sucks. 528 00:19:22,160 --> 00:19:24,240 Like only Effect is like 529 00:19:24,240 --> 00:19:26,320 doing something decent with it. 530 00:19:26,320 --> 00:19:28,720 And I'm like, okay, like take that with a 531 00:19:28,720 --> 00:19:30,480 grain of salt, but 532 00:19:30,480 --> 00:19:32,000 file it away for later. 533 00:19:32,000 --> 00:19:34,640 Kind of go to the Effect docs eventually. 534 00:19:34,640 --> 00:19:37,040 A bit daunting like, all right, what is 535 00:19:37,040 --> 00:19:39,600 this thing on TypeScript? 536 00:19:39,600 --> 00:19:42,080 And eventually sort of through osmosis 537 00:19:42,080 --> 00:19:44,080 and seeing like code snippets of people, 538 00:19:44,080 --> 00:19:44,960 especially like the 539 00:19:44,960 --> 00:19:46,400 Vercel team posts a lot. 540 00:19:46,400 --> 00:19:49,040 Like, okay, this could be nice. 541 00:19:49,040 --> 00:19:51,600 And then looking for sort of some project 542 00:19:51,600 --> 00:19:53,840 to start it with some like 543 00:19:53,840 --> 00:19:56,800 some excuse to write with effects. 544 00:19:56,800 --> 00:19:59,360 And finally, we had some little proof of 545 00:19:59,360 --> 00:20:01,520 concept very isolated from our product 546 00:20:01,520 --> 00:20:03,760 that we were going to build. 547 00:20:03,760 --> 00:20:06,640 It was actually an AI browser agent. 548 00:20:06,640 --> 00:20:10,320 So it was automating browser actions and 549 00:20:10,320 --> 00:20:12,000 computer use with AI. 550 00:20:12,000 --> 00:20:13,440 And I was like, you know what, screw it, 551 00:20:13,440 --> 00:20:15,200 I'm going to use Effect for this. 552 00:20:15,200 --> 00:20:17,520 And that's sort of my first experience. 553 00:20:17,520 --> 00:20:18,640 It was very slow going. 554 00:20:18,640 --> 00:20:20,320 There is definitely a learning curve. 555 00:20:20,320 --> 00:20:22,880 And that's sort of the impetus. 556 00:20:22,880 --> 00:20:25,760 And I guess I can mention sort of to why 557 00:20:25,760 --> 00:20:27,920 I was even ready for 558 00:20:27,920 --> 00:20:29,680 something like Effect. 559 00:20:30,320 --> 00:20:34,320 We like started early on TypeScript, as I 560 00:20:34,320 --> 00:20:36,240 mentioned, because it was just like the 561 00:20:36,240 --> 00:20:37,360 easiest way to get started. 562 00:20:37,360 --> 00:20:39,040 You can move very fast. 563 00:20:39,040 --> 00:20:41,600 And it's very optimized 564 00:20:41,600 --> 00:20:44,400 for like happy path coding. 565 00:20:44,400 --> 00:20:47,440 It's like, you know, those old TVs with 566 00:20:47,440 --> 00:20:49,920 the antennas on top, you like hit it a 567 00:20:49,920 --> 00:20:50,640 bunch of times and 568 00:20:50,640 --> 00:20:52,000 eventually it like starts working. 569 00:20:52,000 --> 00:20:53,760 It's like static and you just hit it. 570 00:20:53,760 --> 00:20:56,080 That's how it feels like coding in 571 00:20:56,080 --> 00:20:59,040 TypeScript or in JavaScript without like 572 00:20:59,040 --> 00:21:01,440 proper result types 573 00:21:01,440 --> 00:21:02,960 and dependency injection. 574 00:21:02,960 --> 00:21:05,040 It's just like, okay, this error happens. 575 00:21:05,040 --> 00:21:06,800 Let me like add some 576 00:21:06,800 --> 00:21:08,240 conditional for that. 577 00:21:08,240 --> 00:21:10,960 And then like eventually like a thousand 578 00:21:10,960 --> 00:21:12,720 times later, you have like something that 579 00:21:12,720 --> 00:21:13,680 works and you're way too 580 00:21:13,680 --> 00:21:15,120 scared to touch it again. 581 00:21:15,120 --> 00:21:16,960 That's how it feels like coding in 582 00:21:16,960 --> 00:21:19,120 vanilla sort of TypeScript, JavaScript. 583 00:21:19,120 --> 00:21:20,240 So I'm like, all right, this 584 00:21:20,240 --> 00:21:21,840 is like just not sustainable. 585 00:21:21,840 --> 00:21:23,760 We have a product where, you know, 586 00:21:23,760 --> 00:21:26,400 everything has five or six different 587 00:21:26,400 --> 00:21:28,960 states it could be in, statuses. 588 00:21:28,960 --> 00:21:32,240 And then every and then there's like, say 589 00:21:32,240 --> 00:21:34,320 20 objects with 20 different statuses. 590 00:21:34,320 --> 00:21:37,040 Like it's just an exponential amount of 591 00:21:37,040 --> 00:21:39,440 different statuses things can be in. 592 00:21:39,440 --> 00:21:40,640 So a lot of antennas 593 00:21:40,640 --> 00:21:42,000 that you need to slap. 594 00:21:43,040 --> 00:21:43,680 Exactly. 595 00:21:43,680 --> 00:21:44,800 That's what I'm saying. 596 00:21:44,800 --> 00:21:46,320 So I'm like, I need 597 00:21:46,320 --> 00:21:47,760 something that can do this better. 598 00:21:47,760 --> 00:21:49,520 And so that's why I was like very 599 00:21:49,520 --> 00:21:51,600 receptive initially to Effect. 600 00:21:51,600 --> 00:21:55,440 So this is one analogy I make quite a bit 601 00:21:55,440 --> 00:21:57,600 of sort of like the sugar high versus 602 00:21:57,600 --> 00:21:59,600 eating vegetables, like 603 00:21:59,600 --> 00:22:01,440 getting building something quickly. 604 00:22:01,440 --> 00:22:02,240 This is like, I mean, 605 00:22:02,240 --> 00:22:04,640 JavaScript is like perfect at that. 606 00:22:04,640 --> 00:22:06,720 But then like when you like run into 607 00:22:06,720 --> 00:22:09,200 those problems where in production you 608 00:22:09,200 --> 00:22:11,440 see undefined is not a function and you 609 00:22:11,440 --> 00:22:12,800 have no clue where 610 00:22:12,800 --> 00:22:13,840 this might even happen. 611 00:22:13,840 --> 00:22:16,560 Maybe the the strike phrase is like, it's 612 00:22:16,560 --> 00:22:19,440 somewhere like some esoteric place and 613 00:22:19,440 --> 00:22:21,760 you're just like completely lost and easy 614 00:22:21,760 --> 00:22:22,880 kind of like dawns on you. 615 00:22:22,880 --> 00:22:24,640 Yes, I should have eaten more vegetables. 616 00:22:24,640 --> 00:22:25,280 I should have built 617 00:22:25,280 --> 00:22:26,560 this in the right way. 618 00:22:26,560 --> 00:22:29,440 But now to kind of like get from place 619 00:22:29,440 --> 00:22:32,000 A to place B where like 620 00:22:32,000 --> 00:22:32,800 you're currently in 621 00:22:32,800 --> 00:22:34,800 undefined is not a function place. 622 00:22:34,800 --> 00:22:36,560 And you're in that place where everything 623 00:22:36,560 --> 00:22:38,080 is nicely modeled like 624 00:22:38,080 --> 00:22:39,840 that can be quite daunting. 625 00:22:39,840 --> 00:22:42,880 So you've taken the plunge with the A.I. 626 00:22:42,880 --> 00:22:45,600 browser automation project. 627 00:22:45,600 --> 00:22:47,280 Can you share a little bit more of like 628 00:22:47,280 --> 00:22:48,560 what that looked like? 629 00:22:48,560 --> 00:22:50,240 What were you trying to did 630 00:22:50,240 --> 00:22:51,360 you build that from scratch? 631 00:22:51,360 --> 00:22:53,760 Did you have like an existing version 632 00:22:53,760 --> 00:22:56,000 that's like has worked in happy path 633 00:22:56,000 --> 00:22:58,720 land, but you now want to make production 634 00:22:58,720 --> 00:23:00,960 ready or like take me through like the 635 00:23:00,960 --> 00:23:02,880 adoption journey of Effect 636 00:23:02,880 --> 00:23:04,160 for that particular project. 637 00:23:04,160 --> 00:23:05,200 And how did you involve 638 00:23:05,200 --> 00:23:06,960 your other team members? 639 00:23:06,960 --> 00:23:08,080 So this project 640 00:23:08,080 --> 00:23:09,680 was very, very isolated. 641 00:23:09,680 --> 00:23:11,440 It was a from scratch product. 642 00:23:11,440 --> 00:23:12,240 We haven't done 643 00:23:12,240 --> 00:23:13,440 anything like that before. 644 00:23:13,440 --> 00:23:16,480 It was it was sort of R and D just kind 645 00:23:16,480 --> 00:23:17,600 of going out seeing 646 00:23:17,600 --> 00:23:18,880 if this would work. 647 00:23:18,880 --> 00:23:21,280 And and actually in hindsight, like that 648 00:23:21,280 --> 00:23:24,560 was probably not the right place to start 649 00:23:24,560 --> 00:23:26,480 with Effect because we were just trying 650 00:23:26,480 --> 00:23:28,400 to build a proof of concept. 651 00:23:28,400 --> 00:23:31,120 But I was like so dead set on on 652 00:23:31,120 --> 00:23:33,280 improving our workflows that I was trying 653 00:23:33,280 --> 00:23:36,320 to use this thing I wanted to use. 654 00:23:36,320 --> 00:23:37,760 You know, startups are great 655 00:23:37,760 --> 00:23:39,680 because you can just kind of choose your 656 00:23:39,680 --> 00:23:40,640 tech stack as you go 657 00:23:40,640 --> 00:23:42,080 along and try things out. 658 00:23:42,080 --> 00:23:44,400 You're not like constrained to other 659 00:23:44,400 --> 00:23:46,160 internal libraries or anything. 660 00:23:46,160 --> 00:23:48,560 So super excited to try it there, even 661 00:23:48,560 --> 00:23:50,560 though it wasn't the best fit. 662 00:23:50,560 --> 00:23:53,600 I was using this library called Stagehand 663 00:23:53,600 --> 00:23:55,360 from Browserbase that kind of 664 00:23:55,360 --> 00:23:57,680 integrates with different providers and 665 00:23:57,680 --> 00:24:00,640 and read sort of the DOM to extract 666 00:24:00,640 --> 00:24:02,320 actionable things in on any 667 00:24:02,320 --> 00:24:04,400 web page and then act upon them. 668 00:24:04,400 --> 00:24:06,560 So I was just kind of like wrapping 669 00:24:06,560 --> 00:24:08,080 functions with Effect and 670 00:24:08,080 --> 00:24:09,440 like seeing how it worked. 671 00:24:09,440 --> 00:24:10,160 Like how does this 672 00:24:10,160 --> 00:24:12,080 dependency injection thing work? 673 00:24:12,080 --> 00:24:13,440 So that's sort of how it got started. 674 00:24:13,440 --> 00:24:15,200 I was actually working on that 675 00:24:15,200 --> 00:24:17,280 project pretty much alone. 676 00:24:17,280 --> 00:24:19,520 And then we built out the proof of 677 00:24:19,520 --> 00:24:21,360 concept where like this can work, we're 678 00:24:21,360 --> 00:24:22,960 going to productize it later. 679 00:24:22,960 --> 00:24:24,080 But I need to go back 680 00:24:24,080 --> 00:24:25,840 to like building product. 681 00:24:25,840 --> 00:24:29,280 So it was sort of a decent amount of time 682 00:24:29,280 --> 00:24:32,800 between that experience and all right, we 683 00:24:32,800 --> 00:24:34,400 need to get this into our product. 684 00:24:34,400 --> 00:24:35,600 And I think the 685 00:24:35,600 --> 00:24:36,960 sort of thing that pushed me 686 00:24:36,960 --> 00:24:39,440 was actually hiring new people. 687 00:24:39,440 --> 00:24:42,720 We have this like mess of of old happy 688 00:24:42,720 --> 00:24:45,440 path code in some places that that needs 689 00:24:45,440 --> 00:24:47,600 to be like more productionized and and 690 00:24:47,600 --> 00:24:49,840 part of it is like an ego thing like, oh, 691 00:24:49,840 --> 00:24:51,840 I have to show this code to someone else. 692 00:24:51,840 --> 00:24:52,960 I have to like someone 693 00:24:52,960 --> 00:24:54,160 else has to work on this. 694 00:24:54,160 --> 00:24:56,080 We need to turn up before 695 00:24:56,080 --> 00:24:57,680 before your friends arrive. 696 00:24:57,680 --> 00:24:59,360 Exactly, exactly. 697 00:24:59,360 --> 00:25:01,760 I'm like, how can I expect like a new 698 00:25:01,760 --> 00:25:05,040 team member or like five members? 699 00:25:05,040 --> 00:25:06,760 I know the feeling this is like you want 700 00:25:06,760 --> 00:25:09,920 someone like it works the right way, but 701 00:25:09,920 --> 00:25:12,080 like you gotta you gotta clean it up now. 702 00:25:12,080 --> 00:25:13,440 Yes. Yeah. 703 00:25:13,440 --> 00:25:16,160 So that that's when we were sort of like, 704 00:25:16,160 --> 00:25:17,440 all right, we need a 705 00:25:17,440 --> 00:25:19,520 better developer experience. 706 00:25:19,520 --> 00:25:22,160 We need everything to be more testable, 707 00:25:22,160 --> 00:25:24,400 everything, all the errors enumerated. 708 00:25:24,400 --> 00:25:27,360 Let's like sit down and sort of if we 709 00:25:27,360 --> 00:25:28,560 were in an ideal world, 710 00:25:28,560 --> 00:25:30,080 what would our stack be? 711 00:25:30,080 --> 00:25:32,640 That still allowed us to move very fast 712 00:25:32,640 --> 00:25:35,520 and allowed people to onboard quickly and 713 00:25:35,520 --> 00:25:36,480 write both front end 714 00:25:36,480 --> 00:25:37,920 and back end very quickly. 715 00:25:37,920 --> 00:25:39,760 And we sort of came up with that answer 716 00:25:39,760 --> 00:25:40,800 and we were like, all 717 00:25:40,800 --> 00:25:42,800 right, let's just go for it. 718 00:25:42,800 --> 00:25:46,640 And just to share sort of like one 719 00:25:46,640 --> 00:25:48,800 anecdote I've heard from a couple of 720 00:25:48,800 --> 00:25:51,040 other companies who are using Effect. 721 00:25:51,040 --> 00:25:53,920 That is it's actually interesting that 722 00:25:53,920 --> 00:25:57,120 they also hired people with no prior 723 00:25:57,120 --> 00:25:58,640 TypeScript experience. 724 00:25:58,640 --> 00:26:00,560 And this was sort of like the first time 725 00:26:00,560 --> 00:26:02,640 really where someone went super deep on 726 00:26:02,640 --> 00:26:04,840 TypeScript and went right into Effect. 727 00:26:04,840 --> 00:26:07,680 I was like me as a Effect team member, 728 00:26:07,680 --> 00:26:09,360 this felt initially kind of 729 00:26:09,360 --> 00:26:12,640 counterintuitive because I came to Effect 730 00:26:12,640 --> 00:26:15,440 because I've done TypeScript for like a 731 00:26:15,440 --> 00:26:17,920 decade, suffering through the 732 00:26:17,920 --> 00:26:19,440 limitations of TypeScript. 733 00:26:19,440 --> 00:26:21,280 And then I saw Effect 734 00:26:21,280 --> 00:26:23,760 as the answer to that. 735 00:26:23,760 --> 00:26:25,680 But I've already learned sort of like all 736 00:26:25,680 --> 00:26:28,880 of like those mishabits of like holding 737 00:26:28,880 --> 00:26:32,080 it in this particular way that like tribe 738 00:26:32,080 --> 00:26:32,960 knowledge has just 739 00:26:32,960 --> 00:26:34,960 been passed down to you. 740 00:26:34,960 --> 00:26:37,120 But people outside of the TypeScript 741 00:26:37,120 --> 00:26:39,440 ecosystem, they've never been burdened 742 00:26:39,440 --> 00:26:43,840 with all of those like crafty experience 743 00:26:43,840 --> 00:26:45,840 like you're using this way not 744 00:26:45,840 --> 00:26:47,920 because it's particularly nice, but just 745 00:26:47,920 --> 00:26:49,840 because you have to do it this way. 746 00:26:49,840 --> 00:26:52,000 But if you're new to TypeScript and new 747 00:26:52,000 --> 00:26:53,680 to Effect, then like you're learning it 748 00:26:53,680 --> 00:26:55,200 the right way right away. 749 00:26:55,200 --> 00:26:57,920 And it's weirdly enough, it's actually 750 00:26:57,920 --> 00:27:00,800 easier to learn Effect if you don't have 751 00:27:00,800 --> 00:27:03,120 that year's worth of like using 752 00:27:03,120 --> 00:27:04,560 TypeScript already in 753 00:27:04,560 --> 00:27:06,400 this particular way. 754 00:27:06,400 --> 00:27:08,800 And so that really resonates now. 755 00:27:08,800 --> 00:27:11,760 Well, it is almost like writing or 756 00:27:11,760 --> 00:27:12,960 reading a new language. 757 00:27:12,960 --> 00:27:16,960 Like you never really use generators in 758 00:27:16,960 --> 00:27:19,680 like your day-to-day TypeScript. 759 00:27:19,680 --> 00:27:20,560 Like you'll never see 760 00:27:20,560 --> 00:27:22,560 that in a React app. 761 00:27:22,560 --> 00:27:24,800 And then the type signatures are just 762 00:27:24,800 --> 00:27:25,840 completely different from 763 00:27:25,840 --> 00:27:28,000 something you write yourself. 764 00:27:28,000 --> 00:27:30,000 Yeah. And I think that's the most useful 765 00:27:30,000 --> 00:27:32,400 mental model that you can give like a 766 00:27:32,400 --> 00:27:33,680 newcomer, being 767 00:27:33,680 --> 00:27:36,400 unburdened from like things before. 768 00:27:36,400 --> 00:27:38,720 Yes, the type signatures, etc. 769 00:27:38,720 --> 00:27:40,320 I think that's helpful to have an 770 00:27:40,320 --> 00:27:41,360 understanding of that. 771 00:27:41,360 --> 00:27:43,360 But I think this is actually one of the 772 00:27:43,360 --> 00:27:46,000 most elegant parts about TypeScript, how 773 00:27:46,000 --> 00:27:47,680 nice the inference works. 774 00:27:47,680 --> 00:27:50,560 But then sort of like to absorb the 775 00:27:51,200 --> 00:27:53,760 mental models of the concepts, etc. 776 00:27:53,760 --> 00:27:55,840 Like if your brain is not constantly 777 00:27:55,840 --> 00:27:57,840 racing of like trying to bind it to a 778 00:27:57,840 --> 00:27:59,680 prior TypeScript thing you've been using, 779 00:27:59,680 --> 00:28:01,360 but just like absorbing 780 00:28:01,360 --> 00:28:04,720 that in a more like free world. 781 00:28:04,720 --> 00:28:06,160 I think that's actually 782 00:28:06,160 --> 00:28:07,440 kind of works super well. 783 00:28:07,440 --> 00:28:09,760 And I think this seems to work super well 784 00:28:09,760 --> 00:28:12,320 for people with prior Rails experience. 785 00:28:12,320 --> 00:28:14,480 And also for that new grad you've 786 00:28:14,480 --> 00:28:16,400 mentioned coming more from a Rust 787 00:28:16,400 --> 00:28:18,640 background, since I think Effect is 788 00:28:18,640 --> 00:28:19,840 actually the closest thing 789 00:28:19,840 --> 00:28:22,160 to Rust we have in TypeScript 790 00:28:22,160 --> 00:28:24,400 in terms of like applying systems 791 00:28:24,400 --> 00:28:26,080 programming mindset 792 00:28:26,080 --> 00:28:28,080 and rigor to TypeScript. 793 00:28:28,080 --> 00:28:29,440 Yeah, I definitely agree. 794 00:28:30,080 --> 00:28:32,640 Going back a bit to writing Haskell, 795 00:28:32,640 --> 00:28:36,560 there's this meme in Haskell where if you 796 00:28:36,560 --> 00:28:37,680 pass the type check, 797 00:28:37,680 --> 00:28:39,040 then your code works. 798 00:28:39,040 --> 00:28:41,120 In vanilla, like JavaScript and 799 00:28:41,120 --> 00:28:43,280 TypeScript, I heard someone joke one time 800 00:28:43,280 --> 00:28:46,640 that TypeScript is a VS Code extension. 801 00:28:46,640 --> 00:28:48,880 And that's really funny to me because 802 00:28:48,880 --> 00:28:50,720 types are just suggestions there. 803 00:28:50,720 --> 00:28:52,800 They're not actually real guarantees. 804 00:28:52,800 --> 00:28:54,480 And Effect is sort of the layer that 805 00:28:54,480 --> 00:28:55,840 makes it feel like my 806 00:28:55,840 --> 00:28:56,800 type is a real thing. 807 00:28:56,800 --> 00:28:58,720 I can actually trust that this thing here 808 00:28:58,720 --> 00:29:00,600 is this thing it says it is. 809 00:29:00,600 --> 00:29:02,880 I go so far in writing 810 00:29:02,880 --> 00:29:04,400 production ready code. 811 00:29:04,400 --> 00:29:07,840 Yeah, so as you then like introduced 812 00:29:07,840 --> 00:29:10,720 Effect into like that first R&D project 813 00:29:10,720 --> 00:29:11,920 and then you brought it 814 00:29:11,920 --> 00:29:13,760 to the rest of the system, 815 00:29:13,760 --> 00:29:16,320 walk me through how you've like, how you 816 00:29:16,320 --> 00:29:18,800 cleaned up before new people joined and 817 00:29:18,800 --> 00:29:21,040 how did Effect step by step. 818 00:29:21,040 --> 00:29:22,160 How did you migrate 819 00:29:22,160 --> 00:29:23,760 your project step by step? 820 00:29:23,760 --> 00:29:26,720 I guess we started with sort of 821 00:29:26,720 --> 00:29:29,440 asking the question, you know, how can we 822 00:29:29,440 --> 00:29:30,960 do this incrementally? 823 00:29:30,960 --> 00:29:33,200 And the first place we sort of used it 824 00:29:33,200 --> 00:29:34,800 was we're using these 825 00:29:35,360 --> 00:29:38,560 NoSQL stores in some places. 826 00:29:38,560 --> 00:29:40,480 And we were like, all right, let's try to 827 00:29:40,480 --> 00:29:43,760 use Schema and just the general Effect 828 00:29:43,760 --> 00:29:46,000 dependency injection to try to more 829 00:29:46,000 --> 00:29:48,000 strongly type some of these NoSQL stores 830 00:29:48,000 --> 00:29:49,120 we're using and see if we 831 00:29:49,120 --> 00:29:50,960 can like parse out our data. 832 00:29:50,960 --> 00:29:54,480 We were using like a lot of places just, 833 00:29:54,480 --> 00:29:56,400 you know, let's get this basically 834 00:29:56,400 --> 00:29:59,520 document object and like just write as 835 00:29:59,520 --> 00:30:01,040 thing at the end of it. 836 00:30:01,040 --> 00:30:03,360 And then it's like that thing and we run 837 00:30:03,360 --> 00:30:05,200 into tons of issues doing that with 838 00:30:05,200 --> 00:30:07,440 that's that happy path coding, you know, 839 00:30:07,440 --> 00:30:09,600 so let's schema parse all of these. 840 00:30:09,600 --> 00:30:10,880 Let's like, be sure we have 841 00:30:10,880 --> 00:30:12,240 the things we say we have. 842 00:30:12,240 --> 00:30:14,720 And that sort of led us to starting to 843 00:30:14,720 --> 00:30:17,840 wrap our like database drivers and our 844 00:30:19,040 --> 00:30:20,400 various domain objects 845 00:30:20,400 --> 00:30:23,280 in schemas and context. 846 00:30:23,280 --> 00:30:25,440 And we're like, you know what, let's just 847 00:30:25,440 --> 00:30:27,440 like, forget that, like forget trying to 848 00:30:27,440 --> 00:30:29,200 like rewrite old stuff. 849 00:30:29,200 --> 00:30:30,720 Let's just come up with our 850 00:30:30,720 --> 00:30:32,720 new stack and and write it. 851 00:30:32,720 --> 00:30:33,680 And so we actually 852 00:30:33,680 --> 00:30:35,360 completely wrote a new server. 853 00:30:35,360 --> 00:30:36,320 It's funny the way it 854 00:30:36,320 --> 00:30:38,800 worked because the other sort of DX 855 00:30:38,800 --> 00:30:41,680 improvement we were looking to get was we 856 00:30:41,680 --> 00:30:43,120 wanted to write less 857 00:30:43,120 --> 00:30:44,240 hooks on the front end. 858 00:30:44,240 --> 00:30:47,760 We wanted Open API specs so that we could just 859 00:30:47,760 --> 00:30:49,520 generate all of our front end code. 860 00:30:49,520 --> 00:30:50,880 So we're looking at different ways to do 861 00:30:50,880 --> 00:30:52,880 that. We eventually found, you know, the 862 00:30:52,880 --> 00:30:54,480 Effect Platform thing. 863 00:30:54,480 --> 00:30:55,840 We're like, all right, let's take our 864 00:30:55,840 --> 00:31:00,400 existing API and write the HTTP 865 00:31:00,400 --> 00:31:02,160 group endpoints that 866 00:31:02,160 --> 00:31:04,240 match like what our API is. 867 00:31:04,240 --> 00:31:06,320 And then we can generate our hooks from 868 00:31:06,320 --> 00:31:08,080 that. And then we go deeper and we 869 00:31:08,080 --> 00:31:09,520 realize, oh, we don't need 870 00:31:09,520 --> 00:31:11,520 to generate anything. 871 00:31:11,520 --> 00:31:14,080 Effect can just like infer the entire 872 00:31:14,080 --> 00:31:16,640 HTTP client. So we'll do that. 873 00:31:16,640 --> 00:31:19,280 And then we're like, you know what, let's 874 00:31:19,280 --> 00:31:21,120 why would we rewrite the old things when 875 00:31:21,120 --> 00:31:23,280 we already have the hooks for them? 876 00:31:23,280 --> 00:31:25,360 Like, let's just actually implement our 877 00:31:25,360 --> 00:31:27,200 new API endpoints 878 00:31:27,200 --> 00:31:29,120 through the effect platform. 879 00:31:29,120 --> 00:31:31,360 And then so we kind of incrementally 880 00:31:31,360 --> 00:31:33,520 after implementing one thing in each 881 00:31:33,520 --> 00:31:35,360 place, we were like, you know what, like, 882 00:31:35,360 --> 00:31:37,280 let's just write the whole thing end to 883 00:31:37,280 --> 00:31:39,920 end in Effect and like start pulling in 884 00:31:39,920 --> 00:31:42,800 like entire flows from the legacy code 885 00:31:42,800 --> 00:31:44,880 into our like sort of new stack. 886 00:31:44,880 --> 00:31:46,800 And that's sort of how we've been going 887 00:31:46,800 --> 00:31:50,480 about it. We use RPC to sort of interface 888 00:31:50,480 --> 00:31:52,000 from our old backend 889 00:31:52,000 --> 00:31:54,160 to our Effect backend. 890 00:31:54,160 --> 00:31:55,680 And that's sort of the bridge where we 891 00:31:55,680 --> 00:31:57,920 need to like make a call for a new 892 00:31:57,920 --> 00:31:59,520 service or a new context. 893 00:31:59,520 --> 00:32:02,080 It is like somewhat of a friction point 894 00:32:02,080 --> 00:32:08,000 to add random effects in non Effect code, 895 00:32:08,000 --> 00:32:10,080 because you do like the run sync, you 896 00:32:10,080 --> 00:32:12,240 have to add the all the dependencies 897 00:32:12,240 --> 00:32:13,280 basically right there. 898 00:32:13,280 --> 00:32:14,880 And then you have to handle all the 899 00:32:14,880 --> 00:32:16,880 errors before you're back in like the 900 00:32:16,880 --> 00:32:18,640 world where you don't know anything. 901 00:32:18,640 --> 00:32:20,720 It's a bit more verbose, but I actually 902 00:32:20,720 --> 00:32:22,480 somewhat prefer it because, 903 00:32:22,480 --> 00:32:25,520 as we sort of adopt more and more that 904 00:32:25,520 --> 00:32:27,920 the sort of scope of that initial run 905 00:32:27,920 --> 00:32:29,040 sync just grows and grows. 906 00:32:29,040 --> 00:32:30,640 And then it's like, okay, now we just 907 00:32:30,640 --> 00:32:32,640 have everything Effect. So that's sort 908 00:32:32,640 --> 00:32:33,440 of how it's been going. 909 00:32:33,440 --> 00:32:35,600 I know that was kind of convoluted, but 910 00:32:35,600 --> 00:32:36,880 hopefully answers the 911 00:32:36,880 --> 00:32:38,880 sort of adoption path for us. 912 00:32:38,880 --> 00:32:41,200 No, it definitely resonates. And I think 913 00:32:41,200 --> 00:32:42,880 it shows sort of, I mean, there, I don't 914 00:32:42,880 --> 00:32:45,280 think there is like the canonical 915 00:32:45,280 --> 00:32:47,440 adoption experience for Effect. 916 00:32:47,440 --> 00:32:50,400 I think it always kind of depends on your 917 00:32:50,400 --> 00:32:52,560 particular pain points. That's typically 918 00:32:52,560 --> 00:32:54,240 what I would recommend to people when 919 00:32:54,240 --> 00:32:55,760 they look into Effect. 920 00:32:55,760 --> 00:32:57,920 I wouldn't recommend them doing one big 921 00:32:57,920 --> 00:33:01,280 bang migration where they don't ship any 922 00:33:01,280 --> 00:33:03,280 new features and just migrate everything. 923 00:33:03,280 --> 00:33:06,240 And that drags on for a very long time. I 924 00:33:06,240 --> 00:33:08,720 would rather recommend doing it 925 00:33:08,720 --> 00:33:11,360 incrementally where you can start it 926 00:33:11,360 --> 00:33:13,440 one day and like in the same day you ship 927 00:33:13,440 --> 00:33:15,600 the first version of Effect. 928 00:33:15,600 --> 00:33:17,840 And that touches just a small part of the 929 00:33:17,840 --> 00:33:19,840 code base, but then like you've gained 930 00:33:19,840 --> 00:33:22,880 sort of that incremental confidence that 931 00:33:22,880 --> 00:33:25,920 you can do that without maybe your 932 00:33:25,920 --> 00:33:27,520 colleagues even knowing. 933 00:33:27,520 --> 00:33:29,760 And step by step for the code that you 934 00:33:29,760 --> 00:33:32,000 own, you can clean it up 935 00:33:32,000 --> 00:33:34,720 and make it more like a home. 936 00:33:34,720 --> 00:33:37,600 And then over time you will realize that 937 00:33:37,600 --> 00:33:39,360 it can absorb a lot of 938 00:33:39,360 --> 00:33:42,160 the manual code you've written before. 939 00:33:42,160 --> 00:33:45,120 And by embracing more parts of Effect, 940 00:33:45,120 --> 00:33:47,200 you can delete a lot of the code. 941 00:33:47,200 --> 00:33:50,480 And then also there is some gluey stuff 942 00:33:50,480 --> 00:33:53,760 that you have while you're still sort of 943 00:33:53,760 --> 00:33:55,920 in the adoption process. 944 00:33:55,920 --> 00:33:58,400 And that's fine. Like that you don't need 945 00:33:58,400 --> 00:34:00,880 to ever move beyond that. 946 00:34:00,880 --> 00:34:03,280 But if you then start adopting even more 947 00:34:03,280 --> 00:34:06,320 Effect, then every like even those parts 948 00:34:06,320 --> 00:34:08,160 become just like seamless, like sort of 949 00:34:08,160 --> 00:34:10,080 the seam becomes seamless. 950 00:34:10,080 --> 00:34:12,000 Where you interact with Effect and then you 951 00:34:12,000 --> 00:34:13,680 have like a single entry point. 952 00:34:13,680 --> 00:34:15,360 And I think this is really like where 953 00:34:15,360 --> 00:34:17,600 when I went first through that, it's 954 00:34:17,600 --> 00:34:20,240 like, oh my gosh, this just felt so good. 955 00:34:20,240 --> 00:34:22,640 Like how everything feels so coherent. 956 00:34:22,640 --> 00:34:24,880 And the last time I had that experience, 957 00:34:24,880 --> 00:34:26,880 I think was like migrating a 958 00:34:26,880 --> 00:34:29,840 giant jQuery code base to react. 959 00:34:29,840 --> 00:34:32,320 And once that was completed and I could 960 00:34:32,320 --> 00:34:35,520 just like delete the jQuery dependency, 961 00:34:35,520 --> 00:34:38,880 that felt sort of like a similar step. 962 00:34:38,880 --> 00:34:39,840 Think back to the 963 00:34:39,840 --> 00:34:43,920 time when you onboarded more of your 964 00:34:43,920 --> 00:34:45,680 engineering colleagues and like you 965 00:34:45,680 --> 00:34:46,960 exposed them to Effect. 966 00:34:46,960 --> 00:34:48,400 how did you take them by the hand 967 00:34:48,400 --> 00:34:50,400 and sort of like showed them around and 968 00:34:50,400 --> 00:34:52,480 help them build an intuition? 969 00:34:52,480 --> 00:34:55,120 I probably sounded a bit like a 970 00:34:55,120 --> 00:34:56,880 crazy person at the time. 971 00:34:56,880 --> 00:34:59,920 If they hadn't heard of effect, I hired 972 00:34:59,920 --> 00:35:03,040 them one with prior TypeScript 973 00:35:03,040 --> 00:35:07,440 experience, one the new grad with none. 974 00:35:07,440 --> 00:35:09,520 And then basically he got a bunch of 975 00:35:09,520 --> 00:35:11,120 TypeScript experience. 976 00:35:11,120 --> 00:35:13,760 And then I sort of come to them and I'm 977 00:35:13,760 --> 00:35:14,880 like, this is amazing. 978 00:35:14,880 --> 00:35:16,720 Like we need to rewrite everything, 979 00:35:16,720 --> 00:35:18,880 basically like everything, you know, 980 00:35:18,880 --> 00:35:19,920 effect works best when 981 00:35:19,920 --> 00:35:21,440 it's like end to end. 982 00:35:21,440 --> 00:35:23,120 It has a way of just like 983 00:35:23,120 --> 00:35:24,640 sucking up more and more. 984 00:35:24,640 --> 00:35:28,000 We joke now that like, oh, another day 985 00:35:28,000 --> 00:35:29,520 like I was like reading through 986 00:35:29,520 --> 00:35:31,840 source code or like deep in the discord. 987 00:35:31,840 --> 00:35:34,240 And I found another primitive that that's 988 00:35:34,240 --> 00:35:35,440 in this library that just 989 00:35:35,440 --> 00:35:36,640 like works perfect for us. 990 00:35:36,640 --> 00:35:39,920 So side note, there's a lot of work to be 991 00:35:39,920 --> 00:35:43,280 done to like get a lot of these effect 992 00:35:43,280 --> 00:35:45,680 primitives and sub packages 993 00:35:45,680 --> 00:35:47,760 to have more documentation and get more 994 00:35:47,760 --> 00:35:48,400 people to know about. 995 00:35:48,400 --> 00:35:50,880 I think this this definitely resonates 996 00:35:50,880 --> 00:35:52,960 because Effect is such a vast 997 00:35:52,960 --> 00:35:56,080 ecosystem and sort of like the along 998 00:35:56,080 --> 00:35:58,000 the lines of the saying like that gift 999 00:35:58,000 --> 00:35:59,120 that keeps on giving. 1000 00:35:59,120 --> 00:36:00,800 But I think it's all about having the 1001 00:36:00,800 --> 00:36:03,120 right mindset since like you could with 1002 00:36:03,120 --> 00:36:04,960 one perspective look at it as like, oh, 1003 00:36:04,960 --> 00:36:06,240 such a vast ecosystem, 1004 00:36:06,240 --> 00:36:07,680 so many things to learn. 1005 00:36:07,680 --> 00:36:09,760 And you could be paralyzed by saying 1006 00:36:09,760 --> 00:36:12,240 like, oh, this is this is too much. I 1007 00:36:12,240 --> 00:36:13,600 won't even get started. 1008 00:36:13,600 --> 00:36:17,040 What I think is much more helpful is like 1009 00:36:17,040 --> 00:36:19,920 picking a few abstractions of Effect, not 1010 00:36:19,920 --> 00:36:22,080 even bothering about the existence of the 1011 00:36:22,080 --> 00:36:24,480 others and like starting to adopt it. 1012 00:36:24,480 --> 00:36:27,120 And like you'll be surprised how great 1013 00:36:27,120 --> 00:36:28,720 this will already feel. 1014 00:36:28,720 --> 00:36:31,280 And then step by step, once you have an 1015 00:36:31,280 --> 00:36:33,120 understanding and an intuition for how 1016 00:36:33,120 --> 00:36:34,480 these pieces work, then 1017 00:36:34,480 --> 00:36:35,840 you're in the right spot to. 1018 00:36:35,840 --> 00:36:38,800 Appreciate the next abstraction and have 1019 00:36:38,800 --> 00:36:42,240 that sort of like improve everything in 1020 00:36:42,240 --> 00:36:44,960 the same way as if we see how other 1021 00:36:44,960 --> 00:36:47,120 ecosystems and languages have have 1022 00:36:47,120 --> 00:36:48,720 evolved over the years, 1023 00:36:48,720 --> 00:36:51,120 something like React hooks. 1024 00:36:51,120 --> 00:36:52,880 They solve particular pain 1025 00:36:52,880 --> 00:36:55,520 points that existed before. 1026 00:36:55,520 --> 00:36:58,240 But if they would have been introduced 1027 00:36:58,240 --> 00:37:00,800 like that in the beginning, maybe that 1028 00:37:00,800 --> 00:37:02,960 would have been too overwhelming. So 1029 00:37:02,960 --> 00:37:04,960 If you fully adopted Effect throughout 1030 00:37:04,960 --> 00:37:07,120 your system, then there's still another 1031 00:37:07,120 --> 00:37:09,840 incremental adoption dimension where you 1032 00:37:09,840 --> 00:37:12,400 can start refactoring things with more 1033 00:37:12,400 --> 00:37:15,600 Effect primitives that you realize, oh, 1034 00:37:15,600 --> 00:37:17,200 this is actually a great fit. 1035 00:37:17,200 --> 00:37:19,360 And I think with that, you can solve 1036 00:37:19,360 --> 00:37:21,280 about modeling your system. 1037 00:37:21,280 --> 00:37:23,200 Right. And once you start sort of 1038 00:37:23,200 --> 00:37:25,840 thinking in Effect, you start having 1039 00:37:25,840 --> 00:37:28,800 these ideas like, oh, what if I wrapped 1040 00:37:28,800 --> 00:37:31,680 this service this way or like abstracted 1041 00:37:31,680 --> 00:37:32,880 this thing this way. 1042 00:37:32,880 --> 00:37:34,880 And you come to realize that the 1043 00:37:34,880 --> 00:37:37,440 ecosystem has already done that. 1044 00:37:38,160 --> 00:37:39,600 And here's the thing that does it. And 1045 00:37:39,600 --> 00:37:41,120 you're like, oh, this is this is actually 1046 00:37:41,120 --> 00:37:43,440 perfect. I don't need to like even write 1047 00:37:43,440 --> 00:37:45,280 anything. I can just use this. 1048 00:37:45,280 --> 00:37:46,800 one example recently was 1049 00:37:46,800 --> 00:37:48,800 just simply caching any 1050 00:37:48,800 --> 00:37:50,720 network call like on Redis. 1051 00:37:50,720 --> 00:37:52,720 And we were like about to write a wrapper 1052 00:37:52,720 --> 00:37:56,080 for just calling an Effect that calls 1053 00:37:56,080 --> 00:37:58,240 some other service and first 1054 00:37:58,240 --> 00:38:00,080 checking the cache in Redis. 1055 00:38:00,080 --> 00:38:02,640 And then we found the tagged requests and 1056 00:38:02,640 --> 00:38:04,240 it's just like, oh, just add a cache key 1057 00:38:04,240 --> 00:38:05,760 and like to provide the 1058 00:38:05,760 --> 00:38:08,240 provide the Redis layer or whatever. 1059 00:38:08,240 --> 00:38:11,200 And it's like, oh, these things already 1060 00:38:11,200 --> 00:38:12,800 exist. It's back to the 1061 00:38:12,800 --> 00:38:15,440 way Rails is sort of very opinionated. 1062 00:38:15,440 --> 00:38:17,360 It's the way that Effect 1063 00:38:17,360 --> 00:38:19,200 has this giant ecosystem. 1064 00:38:19,200 --> 00:38:21,360 It's moving more towards these are the 1065 00:38:21,360 --> 00:38:23,040 ways we should do things and it's almost 1066 00:38:23,040 --> 00:38:25,680 even better because you can implement a 1067 00:38:25,680 --> 00:38:27,840 layer any way you want. 1068 00:38:27,840 --> 00:38:30,880 So it's not as opinionated. It's just you 1069 00:38:30,880 --> 00:38:32,320 have the building blocks for what you 1070 00:38:32,320 --> 00:38:34,000 need. It's like it's like building with 1071 00:38:34,000 --> 00:38:35,200 Legos, but it's almost 1072 00:38:35,200 --> 00:38:37,360 even better than that because 1073 00:38:37,360 --> 00:38:39,600 when you build with Legos, you can't like 1074 00:38:39,600 --> 00:38:42,160 enter in the center and rip out a piece 1075 00:38:42,160 --> 00:38:43,840 and like replace it. 1076 00:38:43,840 --> 00:38:45,440 But with Effect, you can you can rip out 1077 00:38:45,440 --> 00:38:47,600 any layer and replace it with anything 1078 00:38:47,600 --> 00:38:49,440 else at any time as long as it like 1079 00:38:49,440 --> 00:38:50,960 matches the service. 1080 00:38:50,960 --> 00:38:52,640 Exactly. And I think this is sort of 1081 00:38:52,640 --> 00:38:54,640 hinting at the amazing 1082 00:38:54,640 --> 00:38:56,640 concept of composability. 1083 00:38:56,640 --> 00:38:58,800 And like when we say we're talking about 1084 00:38:58,800 --> 00:39:00,160 Legos is like that we 1085 00:39:00,160 --> 00:39:02,080 can click them together. 1086 00:39:02,080 --> 00:39:03,520 We can compose them. 1087 00:39:03,520 --> 00:39:05,120 and there is sort of like the like the 1088 00:39:05,120 --> 00:39:08,560 atomic or subatomic level of like you 1089 00:39:08,560 --> 00:39:10,160 can't get smaller than that. 1090 00:39:10,160 --> 00:39:12,080 But everything else is kind of composed 1091 00:39:12,080 --> 00:39:13,840 out of those. But then you have like 1092 00:39:13,840 --> 00:39:16,720 those like more macro structures that are 1093 00:39:16,720 --> 00:39:18,960 already useful and you can use them. 1094 00:39:18,960 --> 00:39:20,960 But you can also at any time like kind of 1095 00:39:20,960 --> 00:39:23,040 take take those apart and like have the 1096 00:39:23,040 --> 00:39:25,120 little Legos that make the big Lego. 1097 00:39:25,120 --> 00:39:27,120 And I think this is what you rarely have 1098 00:39:27,120 --> 00:39:28,560 in those larger 1099 00:39:28,560 --> 00:39:30,320 ecosystems like Ruby and Rails. 1100 00:39:30,320 --> 00:39:31,920 This is where basically you get the big 1101 00:39:31,920 --> 00:39:33,920 thing. And if it doesn't do what you 1102 00:39:33,920 --> 00:39:35,280 want, well, bad luck. 1103 00:39:35,280 --> 00:39:37,280 You got to roll your own from scratch. 1104 00:39:37,280 --> 00:39:38,800 Whereas with Effect, you can very 1105 00:39:38,800 --> 00:39:41,520 elegantly kind of take those things apart 1106 00:39:41,520 --> 00:39:44,160 and tweak them to your liking. 1107 00:39:44,720 --> 00:39:46,080 I guess back to your 1108 00:39:46,080 --> 00:39:48,080 earlier question about introducing it 1109 00:39:48,080 --> 00:39:50,160 sort of to the to the team and whether 1110 00:39:50,160 --> 00:39:51,840 they were apprehensive. 1111 00:39:51,840 --> 00:39:53,040 They definitely were. 1112 00:39:53,040 --> 00:39:54,880 there's a learning curve and we're always 1113 00:39:54,880 --> 00:39:57,360 just trying to ship as fast as possible. 1114 00:39:57,360 --> 00:40:00,320 We have pressure from customers and 1115 00:40:00,320 --> 00:40:02,560 product guys just to 1116 00:40:02,560 --> 00:40:03,760 ship as fast as possible. 1117 00:40:03,760 --> 00:40:06,080 Like, let's have this out tomorrow. So 1118 00:40:06,080 --> 00:40:08,480 I'm coming to them and saying, learn this 1119 00:40:08,480 --> 00:40:10,720 entirely new way to write TypeScript. 1120 00:40:10,720 --> 00:40:12,880 Trust me, bro, it'll be worth it. 1121 00:40:12,880 --> 00:40:15,040 And it helps to like have some examples I 1122 00:40:15,040 --> 00:40:18,560 can point to. And I think I recall one 1123 00:40:18,560 --> 00:40:20,320 team member of the one actually with 1124 00:40:20,320 --> 00:40:22,240 prior TypeScript experience. 1125 00:40:22,240 --> 00:40:24,640 He he trusted me enough to give it a try. 1126 00:40:24,640 --> 00:40:27,520 And he said it really clicked for him 1127 00:40:27,520 --> 00:40:29,600 when he was writing tests, actually, 1128 00:40:29,600 --> 00:40:32,960 because he like mocked some layer or 1129 00:40:32,960 --> 00:40:35,360 wrote a test layer for for some service. 1130 00:40:35,360 --> 00:40:38,480 And he's like, oh, this is what I can do 1131 00:40:38,480 --> 00:40:41,120 with this and like isolate just certain 1132 00:40:41,120 --> 00:40:42,640 services and tests like 1133 00:40:42,640 --> 00:40:44,160 just the core business logic. 1134 00:40:44,160 --> 00:40:47,200 And it's super, super helpful there. But 1135 00:40:47,200 --> 00:40:48,480 that was the moment that clicked for him. 1136 00:40:48,480 --> 00:40:50,480 I think we've all had sort 1137 00:40:50,480 --> 00:40:52,800 of that that moment, you know, that that 1138 00:40:52,800 --> 00:40:53,600 is just like, oh, this is 1139 00:40:53,600 --> 00:40:55,040 what this is capable of. 1140 00:40:55,040 --> 00:40:56,160 This needs to be sort of 1141 00:40:56,160 --> 00:40:57,680 everywhere on a code base. 1142 00:40:57,680 --> 00:41:00,240 Exactly and I am considering those sort of like 1143 00:41:00,240 --> 00:41:03,360 the superpowers of Effect. And 1144 00:41:03,360 --> 00:41:04,800 you can also turn around those 1145 00:41:04,800 --> 00:41:08,240 superpowers into like super tricky 1146 00:41:08,240 --> 00:41:11,200 problems when you like those being error 1147 00:41:11,200 --> 00:41:13,520 handling those being observability 1148 00:41:13,520 --> 00:41:15,840 testing concurrency, etc. 1149 00:41:15,840 --> 00:41:18,400 Like if your app doesn't need those yet, 1150 00:41:18,400 --> 00:41:20,320 great. And like if you're in that 1151 00:41:20,320 --> 00:41:21,920 position, like Effect 1152 00:41:21,920 --> 00:41:23,600 might make less sense for you. 1153 00:41:23,600 --> 00:41:26,240 But the moment you need those and even 1154 00:41:26,240 --> 00:41:28,800 more if you combine those, then you're in 1155 00:41:28,800 --> 00:41:30,720 a really tricky spot. 1156 00:41:30,720 --> 00:41:32,880 Like, let's say you want to test your 1157 00:41:32,880 --> 00:41:35,360 app. OK, that's hard. But now you want to 1158 00:41:35,360 --> 00:41:38,800 test your app for particularly tricky 1159 00:41:38,800 --> 00:41:41,200 error scenarios that you don't know how 1160 00:41:41,200 --> 00:41:44,400 to kind of like induce into your system. 1161 00:41:44,400 --> 00:41:46,480 like trying to do that without Effect, 1162 00:41:46,480 --> 00:41:47,680 then you basically just spend 1163 00:41:47,680 --> 00:41:49,680 like so much time building your app and 1164 00:41:49,680 --> 00:41:51,680 you spend so much time just building this 1165 00:41:51,680 --> 00:41:54,400 crazy mocking system that like tries to 1166 00:41:54,400 --> 00:41:55,440 simulate everything. 1167 00:41:55,440 --> 00:41:57,680 And you probably feel like, do we really 1168 00:41:57,680 --> 00:42:00,240 need that test? And just this 1169 00:42:00,240 --> 00:42:02,480 incentivizes good software engineering. 1170 00:42:02,480 --> 00:42:03,280 And I think this is 1171 00:42:03,280 --> 00:42:05,040 what Effect really flips. 1172 00:42:05,040 --> 00:42:06,960 There's always a breaking point for 1173 00:42:06,960 --> 00:42:07,520 someone's error 1174 00:42:07,520 --> 00:42:09,520 handling for someone's testing. 1175 00:42:09,520 --> 00:42:11,600 And I think you just need to like 1176 00:42:11,600 --> 00:42:15,040 walk someone to that problem. And then I 1177 00:42:15,040 --> 00:42:17,040 think it like it starts to click. 1178 00:42:17,040 --> 00:42:18,160 Yeah, exactly. This 1179 00:42:18,160 --> 00:42:19,840 composability, interestingly, has 1180 00:42:19,840 --> 00:42:20,800 actually been great 1181 00:42:20,800 --> 00:42:23,440 specifically for payments. 1182 00:42:23,440 --> 00:42:25,040 You have like all these different payment 1183 00:42:25,040 --> 00:42:27,280 rails and and all these different like 1184 00:42:27,280 --> 00:42:29,680 providers and you can just write one 1185 00:42:29,680 --> 00:42:32,160 common interface and the usage 1186 00:42:32,160 --> 00:42:34,240 code doesn't care about what 1187 00:42:34,240 --> 00:42:35,680 rails the payments going on. 1188 00:42:35,680 --> 00:42:37,600 You can change the rails. You can 1189 00:42:37,600 --> 00:42:39,600 dynamically route the payment rails just 1190 00:42:39,600 --> 00:42:42,160 by like providing a different service, 1191 00:42:42,160 --> 00:42:43,280 providing different layers. 1192 00:42:43,280 --> 00:42:45,920 So it's been super interesting to see 1193 00:42:45,920 --> 00:42:46,960 that the patterns we 1194 00:42:46,960 --> 00:42:48,480 can develop this way. 1195 00:42:48,480 --> 00:42:50,640 Can we shift gears a little bit and talk 1196 00:42:50,640 --> 00:42:53,280 a bit more about like the architecture of 1197 00:42:53,280 --> 00:42:57,120 your system? So like if I'm going to 1198 00:42:57,120 --> 00:43:00,800 joinwarp.com, like there's a website and 1199 00:43:00,800 --> 00:43:02,800 probably next to the website, there's 1200 00:43:02,800 --> 00:43:04,800 also some sort of like dashboards that 1201 00:43:04,800 --> 00:43:06,400 your customers are using. 1202 00:43:06,400 --> 00:43:09,280 So that dashboard, I suppose you're using 1203 00:43:09,280 --> 00:43:11,840 something like Next.js for that. So that 1204 00:43:11,840 --> 00:43:14,400 run, there's a front end at minimum. 1205 00:43:14,400 --> 00:43:17,120 But that front end is then talking to 1206 00:43:17,120 --> 00:43:20,560 some sort of like API tier. So we have 1207 00:43:20,560 --> 00:43:23,680 that as well. But I'm curious, what else 1208 00:43:23,680 --> 00:43:26,640 is there? Is there like you have an API 1209 00:43:26,640 --> 00:43:28,080 level, but then you 1210 00:43:28,080 --> 00:43:29,760 probably also have places. 1211 00:43:29,760 --> 00:43:31,680 Maybe you have cron jobs. Maybe you have 1212 00:43:31,680 --> 00:43:33,200 like other things you've mentioned like 1213 00:43:33,200 --> 00:43:37,840 that AI agent browser setup, etc. So I'm 1214 00:43:37,840 --> 00:43:39,120 not sure whether you're thinking about 1215 00:43:39,120 --> 00:43:39,760 this as like 1216 00:43:39,760 --> 00:43:43,120 microservices or majestic monolith. 1217 00:43:43,120 --> 00:43:45,200 But like maybe you can explain a little 1218 00:43:45,200 --> 00:43:48,480 bit of like how that imagine we're at a 1219 00:43:48,480 --> 00:43:50,240 whiteboard and you're drawing the system 1220 00:43:50,240 --> 00:43:51,600 architecture. Like maybe 1221 00:43:51,600 --> 00:43:53,120 you can do that in words. 1222 00:43:53,120 --> 00:43:56,080 We definitely prefer to keep things 1223 00:43:56,080 --> 00:43:58,480 as simple as possible. We're a team of 1224 00:43:58,480 --> 00:44:01,200 four and we don't have high requirements 1225 00:44:01,200 --> 00:44:04,640 for like latency on requests. 1226 00:44:04,640 --> 00:44:08,000 We have requirements for reliability 1227 00:44:08,000 --> 00:44:09,440 being being payment software. 1228 00:44:09,440 --> 00:44:11,760 another thing is we don't have a high 1229 00:44:11,760 --> 00:44:15,920 total number of like daily active users. 1230 00:44:15,920 --> 00:44:18,800 So our compute requirements are actually 1231 00:44:18,800 --> 00:44:21,200 interestingly somewhat low. That being 1232 00:44:21,200 --> 00:44:24,080 said, the way we started was on Next.js 1233 00:44:24,080 --> 00:44:26,240 and almost everything was 1234 00:44:26,240 --> 00:44:28,400 just in a Next.js API route. 1235 00:44:28,400 --> 00:44:30,480 That grew honestly way past its 1236 00:44:30,480 --> 00:44:32,720 usefulness, but that's the way things go 1237 00:44:32,720 --> 00:44:35,120 at startups. You just pick a stack that's 1238 00:44:35,120 --> 00:44:36,800 been the theme of this whole interview. 1239 00:44:36,800 --> 00:44:39,040 You pick a stack, you go too far with it. 1240 00:44:39,040 --> 00:44:40,960 It doesn't meet your needs. 1241 00:44:40,960 --> 00:44:43,760 And also just as a side note, it's almost 1242 00:44:43,760 --> 00:44:46,800 diametrically opposed to what you've been 1243 00:44:46,800 --> 00:44:49,280 building before with the social network 1244 00:44:49,280 --> 00:44:52,160 where like it's like the latency really 1245 00:44:52,160 --> 00:44:55,040 matters, etc. And like it now you've like 1246 00:44:55,040 --> 00:44:56,640 completely flipped all of 1247 00:44:56,640 --> 00:44:58,800 your like tech trade offs. 1248 00:44:58,800 --> 00:45:00,400 Yeah, that one was fun to build because 1249 00:45:00,400 --> 00:45:02,960 it was tons of caching, streaming, like 1250 00:45:02,960 --> 00:45:05,920 all the stuff that needed to scale very 1251 00:45:05,920 --> 00:45:09,360 quickly when there was a usage upburst 1252 00:45:09,360 --> 00:45:10,800 for whatever reason. 1253 00:45:10,800 --> 00:45:13,520 And now our usage patterns are a lot more 1254 00:45:13,520 --> 00:45:15,680 predictable if you think about it. People 1255 00:45:15,680 --> 00:45:17,600 log in to see their paycheck twice a 1256 00:45:17,600 --> 00:45:19,840 month. We just need to like get that part 1257 00:45:19,840 --> 00:45:21,280 right and then make sure 1258 00:45:21,280 --> 00:45:22,480 all the payments are right. 1259 00:45:22,480 --> 00:45:24,560 So I guess to say more about the 1260 00:45:24,560 --> 00:45:26,480 infrastructure, we're running containers. 1261 00:45:26,480 --> 00:45:28,320 We have one monolith pretty much that 1262 00:45:28,320 --> 00:45:33,920 handles all of our API routes, jobs, even 1263 00:45:33,920 --> 00:45:36,640 like yeah asynchronous jobs, cron jobs 1264 00:45:36,640 --> 00:45:39,200 and that's our effect platform API. 1265 00:45:39,200 --> 00:45:40,640 And then we do have like separate 1266 00:45:40,640 --> 00:45:43,680 services, for example, for the browser 1267 00:45:43,680 --> 00:45:46,320 agent thing, we have a slack bot that's 1268 00:45:46,320 --> 00:45:47,760 separate. That's like sufficiently 1269 00:45:47,760 --> 00:45:50,960 different to necessitate another service. 1270 00:45:50,960 --> 00:45:53,200 But all that to say pretty simple, we 1271 00:45:53,200 --> 00:45:55,200 like to work in a monolith. 1272 00:45:55,200 --> 00:45:57,440 And is that monolith currently deployed 1273 00:45:57,440 --> 00:46:00,240 in a single place or are you already 1274 00:46:00,240 --> 00:46:02,320 running that in a distributed setting 1275 00:46:02,320 --> 00:46:04,080 that maybe you have one running in the 1276 00:46:04,080 --> 00:46:05,520 East Coast, one running on the West 1277 00:46:05,520 --> 00:46:08,640 Coast, wherever your customers are? 1278 00:46:08,640 --> 00:46:10,400 Yeah, we just have one running on the 1279 00:46:10,400 --> 00:46:12,960 East Coast that's next to our database. 1280 00:46:12,960 --> 00:46:15,680 Got it. So maybe at some point, there 1281 00:46:15,680 --> 00:46:18,960 might be the day on the horizon where you 1282 00:46:18,960 --> 00:46:21,440 need to like go multi data center, which 1283 00:46:21,440 --> 00:46:25,040 I'm sure will upend a whole bunch of like 1284 00:46:25,040 --> 00:46:26,640 so far trade offs you 1285 00:46:26,640 --> 00:46:27,920 didn't have to make yet. 1286 00:46:27,920 --> 00:46:30,640 But this is another thing where maybe 1287 00:46:30,640 --> 00:46:34,000 I'll invite you back to the podcast to 1288 00:46:34,000 --> 00:46:36,800 see whether Effect Workflows could help 1289 00:46:36,800 --> 00:46:37,920 with that. Since this is 1290 00:46:37,920 --> 00:46:39,440 exactly what this is about. 1291 00:46:39,440 --> 00:46:41,840 Kind of like to like going with a 1292 00:46:41,840 --> 00:46:44,560 monolith and just deploying that monolith 1293 00:46:44,560 --> 00:46:47,520 to a single place is the simplest you can 1294 00:46:47,520 --> 00:46:49,920 go with. And I love simplicity. 1295 00:46:49,920 --> 00:46:52,080 But at some point, you might need to go 1296 00:46:52,080 --> 00:46:54,720 beyond that and then to still go for 1297 00:46:54,720 --> 00:46:57,280 maximum simplicity. I think this is where 1298 00:46:57,280 --> 00:47:00,240 Effect will be another gift will be 1299 00:47:00,240 --> 00:47:02,480 receiving from Effects is like Effect 1300 00:47:02,480 --> 00:47:04,560 workflows that try to help 1301 00:47:04,560 --> 00:47:05,920 exactly with the sort of thing. 1302 00:47:05,920 --> 00:47:07,520 But we'll talk more about that in the 1303 00:47:07,520 --> 00:47:09,120 future. Yeah, I hope to make 1304 00:47:09,120 --> 00:47:11,440 it there. I hope to have these problems. 1305 00:47:11,440 --> 00:47:13,120 but maybe going back to that, so 1306 00:47:13,120 --> 00:47:14,960 you mentioned like you're running this 1307 00:47:14,960 --> 00:47:16,880 right now in a container. You mentioned 1308 00:47:16,880 --> 00:47:19,280 that you run some sort of like jobs, 1309 00:47:19,280 --> 00:47:21,200 background processes, etc. 1310 00:47:21,200 --> 00:47:23,600 How do you orchestrate that right now? Is 1311 00:47:23,600 --> 00:47:25,760 that still mostly driven through what are 1312 00:47:25,760 --> 00:47:28,160 the entry points are those entry points 1313 00:47:28,160 --> 00:47:29,920 are requests. Those entry 1314 00:47:29,920 --> 00:47:32,640 points might be cron jobs. 1315 00:47:32,640 --> 00:47:34,960 Anything else that sort of serves as an 1316 00:47:34,960 --> 00:47:37,680 entry point to kick off code that runs? 1317 00:47:37,680 --> 00:47:41,200 Yeah, so we have the HTTP like sort of 1318 00:47:41,200 --> 00:47:43,480 restful API that that kicks off code 1319 00:47:43,480 --> 00:47:43,920 where we actually wrapped. 1320 00:47:43,920 --> 00:47:46,720 We actually wrapped BullMQ with 1321 00:47:46,720 --> 00:47:48,720 some Effect helpers so that's the 1322 00:47:48,720 --> 00:47:51,200 way we're kind of queuing up jobs and 1323 00:47:51,200 --> 00:47:53,120 processing them. We're still running that 1324 00:47:53,120 --> 00:47:54,480 just on one server, all the 1325 00:47:54,480 --> 00:47:56,000 handlers. We could distribute that 1326 00:47:56,000 --> 00:47:56,880 somewhat simply. 1327 00:47:56,880 --> 00:47:59,200 And then yeah, there's also RPC endpoints 1328 00:47:59,200 --> 00:48:01,440 just for calling more service type 1329 00:48:01,440 --> 00:48:04,000 functions from from our legacy back end. 1330 00:48:04,000 --> 00:48:05,280 that's pretty much it. 1331 00:48:05,280 --> 00:48:08,480 I guess to mention cron we can 1332 00:48:08,480 --> 00:48:12,320 queue like cron type jobs in in our 1333 00:48:12,320 --> 00:48:14,640 BullMQ so we're not officially using like 1334 00:48:14,640 --> 00:48:16,480 cron jobs, but it's it's 1335 00:48:16,480 --> 00:48:18,560 you know daily scheduled jobs. 1336 00:48:18,560 --> 00:48:20,880 Got it. Can you share a bit more about 1337 00:48:20,880 --> 00:48:23,920 sort of like the DevOps lifecycle things 1338 00:48:23,920 --> 00:48:27,040 that that you're dealing with and how 1339 00:48:27,040 --> 00:48:29,040 you're using Effect for that 1340 00:48:29,040 --> 00:48:30,480 part of the picture as well? 1341 00:48:30,480 --> 00:48:31,440 One thing I didn't mention 1342 00:48:31,440 --> 00:48:34,560 is that we're using bun in in our 1343 00:48:34,560 --> 00:48:36,160 container runtime. So that's 1344 00:48:36,160 --> 00:48:38,080 been fun effect pretty much instructs 1345 00:48:38,080 --> 00:48:40,880 that away. So it's it's no different than 1346 00:48:40,880 --> 00:48:43,120 using it in node other than we get to say 1347 00:48:43,120 --> 00:48:44,560 we're running on bun and and 1348 00:48:44,560 --> 00:48:46,720 it's it's fun a bit faster. 1349 00:48:46,720 --> 00:48:47,440 There are some bun 1350 00:48:47,440 --> 00:48:48,880 features that that we like. 1351 00:48:48,880 --> 00:48:50,960 Yeah, I guess we're we're fully automated 1352 00:48:50,960 --> 00:48:53,280 like deployments. We just you know, build 1353 00:48:53,280 --> 00:48:55,920 the container in a GitHub action with bun 1354 00:48:55,920 --> 00:48:58,800 and PNPM and then push it to our 1355 00:48:58,800 --> 00:49:01,840 like ECS cluster on running on AWS. 1356 00:49:01,840 --> 00:49:03,520 Got it. Speaking of sort of like the 1357 00:49:03,520 --> 00:49:06,160 monolith and bun approach a friend of 1358 00:49:06,160 --> 00:49:10,160 mine has been working at Midjourney and 1359 00:49:10,160 --> 00:49:12,160 I think Midjourney actually has also 1360 00:49:12,160 --> 00:49:14,720 embraced the sort of architecture and 1361 00:49:14,720 --> 00:49:17,200 deployment architecture where they have 1362 00:49:17,200 --> 00:49:20,240 like one big machine where they're 1363 00:49:20,240 --> 00:49:22,560 running their monolith through bun and I 1364 00:49:22,560 --> 00:49:24,560 think they call it the mother bun. 1365 00:49:24,560 --> 00:49:27,520 So just a little fun fun side note here. 1366 00:49:27,520 --> 00:49:30,160 But speaking of Midjourney and therefore 1367 00:49:30,160 --> 00:49:32,880 AI, you mentioned initially that one 1368 00:49:32,880 --> 00:49:35,120 theme that you're currently also spending 1369 00:49:35,120 --> 00:49:37,840 quite a lot of time on is automating 1370 00:49:37,840 --> 00:49:41,600 things further with AI. So maybe we can 1371 00:49:41,600 --> 00:49:43,440 pivot the conversation a bit in that 1372 00:49:43,440 --> 00:49:45,280 direction. Do you want to give an 1373 00:49:45,280 --> 00:49:47,360 overview of like what that means and then 1374 00:49:47,360 --> 00:49:49,040 diving closer into that? 1375 00:49:49,040 --> 00:49:52,160 Definitely. The way I think  about AI and our product 1376 00:49:52,160 --> 00:49:54,160 is sort of two 1377 00:49:54,160 --> 00:49:56,080 separate kind of verticals. 1378 00:49:56,080 --> 00:49:59,280 We are a payments product we, you know, 1379 00:49:59,280 --> 00:50:01,760 remit taxes, we have to know all the 1380 00:50:01,760 --> 00:50:04,080 taxes and we have to get those right. So 1381 00:50:04,080 --> 00:50:06,240 you think that plus AI is probably a 1382 00:50:06,240 --> 00:50:07,680 recipe for disaster. 1383 00:50:07,680 --> 00:50:08,880 But I think there are there are sort of 1384 00:50:08,880 --> 00:50:11,360 two ways we we think about it. One is the 1385 00:50:11,360 --> 00:50:13,760 customer facing way, which is, you know, 1386 00:50:13,760 --> 00:50:14,960 we've built this wonderful 1387 00:50:14,960 --> 00:50:17,440 platform for people to use. 1388 00:50:17,440 --> 00:50:21,120 How can we make AI to one make that 1389 00:50:21,120 --> 00:50:24,560 easier to use and two, I think one common 1390 00:50:24,560 --> 00:50:26,480 theme that's going on right now is just 1391 00:50:26,480 --> 00:50:28,640 meeting customers where they are, which, 1392 00:50:28,640 --> 00:50:30,800 for many means building slack bot. 1393 00:50:30,800 --> 00:50:34,080 Chat interfaces, etc. So how can AI 1394 00:50:34,080 --> 00:50:36,000 actually be the bridge between our end 1395 00:50:36,000 --> 00:50:38,560 users and our products in a way that's 1396 00:50:38,560 --> 00:50:41,120 like very safe and human in the loop and 1397 00:50:41,120 --> 00:50:43,280 you know we're not triggering payments 1398 00:50:43,280 --> 00:50:45,120 for people that are that are approved, 1399 00:50:45,120 --> 00:50:46,480 but we can actually use our 1400 00:50:46,480 --> 00:50:48,400 product with natural language. 1401 00:50:48,400 --> 00:50:50,160 And I guess when I say AI here, I'm 1402 00:50:50,160 --> 00:50:52,240 talking about LLMs specifically. 1403 00:50:52,240 --> 00:50:54,160 And then sort of the other way we're 1404 00:50:54,160 --> 00:50:56,800 thinking about it is automating process 1405 00:50:56,800 --> 00:50:59,920 heavy workflows sort of on the back end, 1406 00:50:59,920 --> 00:51:01,680 which is, you know, what if an 1407 00:51:01,680 --> 00:51:05,680 AI can file taxes, what if an AI can like 1408 00:51:05,680 --> 00:51:09,520 read your tax reports and, you know, 1409 00:51:09,520 --> 00:51:12,880 generate insights and and go and and sort 1410 00:51:12,880 --> 00:51:14,640 of do the things that need to be done. 1411 00:51:14,640 --> 00:51:16,640 And these are things that, in 1412 00:51:16,640 --> 00:51:20,240 the pre 2023 era, you might have needed 1413 00:51:20,240 --> 00:51:22,960 an army of employees or contractors to 1414 00:51:22,960 --> 00:51:25,520 like, go through these, you know, 1415 00:51:25,520 --> 00:51:27,440 convoluted .gov websites and 1416 00:51:27,440 --> 00:51:28,960 click through these things. 1417 00:51:28,960 --> 00:51:30,480 But we're sort of asking the question, 1418 00:51:30,480 --> 00:51:31,920 you know, what if, what if a person 1419 00:51:31,920 --> 00:51:33,440 doesn't actually have to do that. 1420 00:51:33,440 --> 00:51:34,960 So those are sort of the different 1421 00:51:34,960 --> 00:51:36,560 ways we're sort of thinking about it. 1422 00:51:36,560 --> 00:51:39,440 Got it. That that is super helpful. So in 1423 00:51:39,440 --> 00:51:41,120 regards to the the latter, 1424 00:51:41,120 --> 00:51:42,240 automating those 1425 00:51:42,240 --> 00:51:44,080 various like manual processes. 1426 00:51:44,080 --> 00:51:46,560 When you mentioned that currently work is 1427 00:51:46,560 --> 00:51:49,440 16 people, 4 of them being engineers, I 1428 00:51:49,440 --> 00:51:52,080 suppose, like a big chunk of the other 1429 00:51:52,080 --> 00:51:55,600 employees are still doing that sort of 1430 00:51:55,600 --> 00:51:57,600 like making the entire 1431 00:51:57,600 --> 00:52:00,240 thing happen behind the scenes. 1432 00:52:00,240 --> 00:52:02,880 But there is, as you scale up your 1433 00:52:02,880 --> 00:52:05,200 customers, you would need to scale up the 1434 00:52:05,200 --> 00:52:08,560 employees. So you want to like automate 1435 00:52:08,560 --> 00:52:10,240 as much as possible. 1436 00:52:10,240 --> 00:52:12,880 And this is where AI like is helpful. 1437 00:52:12,880 --> 00:52:16,000 Yeah, exactly. The economics of it, I 1438 00:52:16,000 --> 00:52:18,640 mean, is that, you know, we have to scale 1439 00:52:18,640 --> 00:52:20,960 our team somewhat linearly with the 1440 00:52:20,960 --> 00:52:22,320 number of customers we have. 1441 00:52:22,320 --> 00:52:25,120 But with AI, you can presumably pay 1442 00:52:25,120 --> 00:52:27,840 pennies for things that you'd have to 1443 00:52:27,840 --> 00:52:29,200 hire a bunch of people for. 1444 00:52:29,200 --> 00:52:30,480 So we're exploring 1445 00:52:30,480 --> 00:52:32,560 different different ways to do that. 1446 00:52:32,560 --> 00:52:34,080 Got it. And now it's like a great 1447 00:52:34,080 --> 00:52:36,640 opportunities for your existing employees 1448 00:52:36,640 --> 00:52:38,800 who've been doing that so far, like help 1449 00:52:38,800 --> 00:52:40,160 in that automation process 1450 00:52:40,160 --> 00:52:42,240 as they're the domain experts. 1451 00:52:42,240 --> 00:52:44,800 And yeah, that sounds fascinating. It 1452 00:52:44,800 --> 00:52:46,960 gets us full circle back to the initial 1453 00:52:46,960 --> 00:52:48,880 point of like we can be thankful to 1454 00:52:48,880 --> 00:52:50,400 you like doing that. 1455 00:52:50,400 --> 00:52:53,600 And now it's actually super fun for the 1456 00:52:53,600 --> 00:52:55,440 people who were the domain experts that 1457 00:52:55,440 --> 00:52:57,440 they can now start being 1458 00:52:57,440 --> 00:52:59,760 included in the automation process. 1459 00:52:59,760 --> 00:53:02,080 So that makes a lot of sense. And I think 1460 00:53:02,080 --> 00:53:05,280 that's a whole different topic of what it 1461 00:53:05,280 --> 00:53:08,080 means to do that automation. 1462 00:53:08,080 --> 00:53:10,080 And there's like a lot of like just from 1463 00:53:10,080 --> 00:53:13,800 my own experiments and like work on AI 1464 00:53:13,800 --> 00:53:16,160 driven things. There's like so many 1465 00:53:16,160 --> 00:53:17,600 things that can just go wrong. 1466 00:53:17,600 --> 00:53:19,520 I mean, if you're just like as a human 1467 00:53:19,520 --> 00:53:22,800 navigating websites, like things just 1468 00:53:22,800 --> 00:53:24,800 rarely are going 1469 00:53:24,800 --> 00:53:26,160 as they as they should. 1470 00:53:26,160 --> 00:53:28,240 There's like a downtime. There's like 1471 00:53:28,240 --> 00:53:30,880 whatever like a page just 1472 00:53:30,880 --> 00:53:33,600 reloads without you triggering it. 1473 00:53:33,600 --> 00:53:36,320 There's just so many unpredictable things 1474 00:53:36,320 --> 00:53:40,400 or sometimes they don't use HTML selects 1475 00:53:40,400 --> 00:53:42,960 and you know, reading the DOM, you have 1476 00:53:42,960 --> 00:53:44,400 no idea what to like there are 1477 00:53:44,400 --> 00:53:45,920 accessibility issues that 1478 00:53:45,920 --> 00:53:47,680 that an AI can run into. 1479 00:53:47,680 --> 00:53:50,320 There are so so many things. Yeah. So 1480 00:53:50,320 --> 00:53:51,760 this is it's whole 1481 00:53:51,760 --> 00:53:53,120 different kind of words. 1482 00:53:53,120 --> 00:53:55,200 And like honestly, it becomes like a 1483 00:53:55,200 --> 00:53:58,160 whole different profession of like how to 1484 00:53:58,160 --> 00:54:01,440 how to build systems that like can 1485 00:54:01,440 --> 00:54:02,720 automate these things 1486 00:54:02,720 --> 00:54:04,080 that we have in mind. 1487 00:54:04,080 --> 00:54:06,720 But maybe we focus a little bit more on 1488 00:54:06,720 --> 00:54:09,360 the first part you've mentioned where 1489 00:54:09,360 --> 00:54:11,520 you're like where you meet people where 1490 00:54:11,520 --> 00:54:13,840 they are and like expose sort of like 1491 00:54:13,840 --> 00:54:16,560 those more AI native interfaces to them. 1492 00:54:16,560 --> 00:54:18,800 Are you also building those systems with 1493 00:54:18,800 --> 00:54:20,960 Effect and maybe you can also share a 1494 00:54:20,960 --> 00:54:22,960 bit more of like how those are built 1495 00:54:22,960 --> 00:54:24,320 which of the Effect's 1496 00:54:24,320 --> 00:54:26,000 primitives they're using. 1497 00:54:26,000 --> 00:54:28,960 Yeah, we are using Effect for it. 1498 00:54:28,960 --> 00:54:30,160 We haven't actually 1499 00:54:30,160 --> 00:54:32,640 used the Effect AI thing. 1500 00:54:32,640 --> 00:54:35,760 I have wanted to use that but I guess our 1501 00:54:35,760 --> 00:54:37,680 team member who's doing it is it like I 1502 00:54:37,680 --> 00:54:39,760 said came from the Python world and it's 1503 00:54:39,760 --> 00:54:42,560 very like fast iteration cycles. 1504 00:54:42,560 --> 00:54:45,600 He was kind of fighting the learning 1505 00:54:45,600 --> 00:54:46,400 curve of learning 1506 00:54:46,400 --> 00:54:48,400 effect versus trying to get features out. 1507 00:54:48,400 --> 00:54:51,120 So we haven't explored that specific 1508 00:54:51,120 --> 00:54:53,920 primitive too much. We are using Effect 1509 00:54:53,920 --> 00:54:55,760 sort of in the same way we've used 1510 00:54:55,760 --> 00:54:58,160 we've just wrapped like the providers. 1511 00:54:58,160 --> 00:55:00,880 we're only using one LLM right now 1512 00:55:00,880 --> 00:55:02,560 which which I'll refrain from 1513 00:55:02,560 --> 00:55:04,480 naming. But you could guess. 1514 00:55:04,480 --> 00:55:06,800 Yeah, sort of the same way just just end 1515 00:55:06,800 --> 00:55:10,480 to end like wrapping workflows. There's 1516 00:55:10,480 --> 00:55:12,320 very much human in the loop sort of 1517 00:55:12,320 --> 00:55:14,560 things and we have to ask on a lot of 1518 00:55:14,560 --> 00:55:16,080 workflows like for 1519 00:55:16,080 --> 00:55:17,280 confirmation from the user. 1520 00:55:17,280 --> 00:55:19,280 So there's these like very stateful 1521 00:55:19,280 --> 00:55:22,880 flows. So we're actually using sort of 1522 00:55:22,880 --> 00:55:25,520 Effect and TypeScript more generally to 1523 00:55:25,520 --> 00:55:26,960 kind of pick up these workflows where 1524 00:55:26,960 --> 00:55:28,720 they left off and write 1525 00:55:28,720 --> 00:55:30,560 like these stateful flows. 1526 00:55:30,560 --> 00:55:32,640 Got it. That makes a lot of sense. And I 1527 00:55:32,640 --> 00:55:34,880 think it's always like a good judgment 1528 00:55:34,880 --> 00:55:38,480 call to see like hey is this is using 1529 00:55:38,480 --> 00:55:41,200 Effect here like today. Is that solving 1530 00:55:41,200 --> 00:55:44,240 a problem we have or would it potentially 1531 00:55:44,240 --> 00:55:48,400 if I pay that learning cost is worthwhile 1532 00:55:48,400 --> 00:55:50,400 paying that today or should 1533 00:55:50,400 --> 00:55:53,440 I pay it in two months from now when I 1534 00:55:53,440 --> 00:55:55,040 have more acute problems. 1535 00:55:55,040 --> 00:55:56,480 that's always a good 1536 00:55:56,480 --> 00:55:59,120 question to ask and I think that starts 1537 00:55:59,120 --> 00:56:02,720 adding up but rest assured that Effect AI 1538 00:56:02,720 --> 00:56:06,720 has become a lot more polished and I use 1539 00:56:06,720 --> 00:56:08,560 it relentlessly in many 1540 00:56:08,560 --> 00:56:11,520 projects and it's an absolute joy. 1541 00:56:11,520 --> 00:56:14,720 Oh, I'll have to explore it deeper. Yeah, 1542 00:56:14,720 --> 00:56:16,560 like you mentioned, it's hard to justify 1543 00:56:16,560 --> 00:56:19,760 when one the person doesn't know it and 1544 00:56:19,760 --> 00:56:22,800 two when you don't know what your data is 1545 00:56:22,800 --> 00:56:24,720 going to end up looking like and you 1546 00:56:24,720 --> 00:56:25,520 don't know if the 1547 00:56:25,520 --> 00:56:27,600 project will even survive. 1548 00:56:27,600 --> 00:56:29,360 So there is there is this. You know back 1549 00:56:29,360 --> 00:56:31,120 and forth between like is the 1550 00:56:31,120 --> 00:56:32,080 learning curve worth it 1551 00:56:32,080 --> 00:56:34,800 versus I need to ship features. 1552 00:56:34,800 --> 00:56:37,040 Yeah, so maybe looking a little bit more 1553 00:56:37,040 --> 00:56:41,440 into the future. What are your hopes for 1554 00:56:41,440 --> 00:56:44,720 Effect as an ecosystem and what are your 1555 00:56:44,720 --> 00:56:46,320 plans for Warp from 1556 00:56:46,320 --> 00:56:47,760 engineering perspective. 1557 00:56:47,760 --> 00:56:50,160 Good question. I think I'm I've been 1558 00:56:50,160 --> 00:56:52,080 super happy with the stuff we're 1559 00:56:52,080 --> 00:56:53,600 building. It's been like such a 1560 00:56:53,600 --> 00:56:56,160 refreshing experience building the new 1561 00:56:56,160 --> 00:56:58,640 way that whenever I'm like digging deep 1562 00:56:58,640 --> 00:57:02,880 into legacy code. I'm just like upset in general. 1563 00:57:02,880 --> 00:57:05,040 I'm like ashamed of what I've what I've 1564 00:57:05,040 --> 00:57:08,240 written in the past in some ways. So I'm 1565 00:57:08,240 --> 00:57:10,880 super excited. I'm very excited for a lot 1566 00:57:10,880 --> 00:57:12,240 of the things that are marked 1567 00:57:12,240 --> 00:57:14,240 experimental to just be like 1568 00:57:14,240 --> 00:57:16,160 production ready, so to speak. 1569 00:57:16,160 --> 00:57:18,240 Super excited for more things like the AI 1570 00:57:18,240 --> 00:57:20,800 package and and also like I mentioned. I 1571 00:57:20,800 --> 00:57:23,360 think there's so many opportunities to 1572 00:57:23,360 --> 00:57:26,720 get a lot of the ecosystem in better like 1573 00:57:26,720 --> 00:57:28,960 documentation states. There are so many 1574 00:57:28,960 --> 00:57:32,080 like discord threads with gems that you 1575 00:57:32,080 --> 00:57:34,240 know the the person coming to Effect for 1576 00:57:34,240 --> 00:57:36,160 the first time will never see for the 1577 00:57:36,160 --> 00:57:37,200 thing that they need 1578 00:57:37,200 --> 00:57:40,400 I think I'm excited for it to be in more 1579 00:57:40,400 --> 00:57:44,320 of the LLM training data. I've seen a lot of. 1580 00:57:44,320 --> 00:57:45,760 They spin their wheels because 1581 00:57:45,760 --> 00:57:47,520 they don't know the syntax or 1582 00:57:47,520 --> 00:57:49,040 whatever. They don't know a specific 1583 00:57:49,040 --> 00:57:51,600 diagnostic error. The LSP extension helps 1584 00:57:51,600 --> 00:57:54,000 a lot with that. So really, I'm just 1585 00:57:54,000 --> 00:57:55,920 excited for it to hit the mainstream more 1586 00:57:55,920 --> 00:57:58,480 and and I'd like to you know do anything 1587 00:57:58,480 --> 00:57:59,680 I can to support that. 1588 00:57:59,680 --> 00:58:00,240 There's a lot of 1589 00:58:00,240 --> 00:58:02,240 opportunities for educational content. I 1590 00:58:02,240 --> 00:58:04,560 see a lot of your team to their credit. 1591 00:58:04,560 --> 00:58:07,040 posting like animations and 1592 00:58:07,040 --> 00:58:09,040 different things. So I'm super excited 1593 00:58:09,040 --> 00:58:11,520 for the direction it's going and I think 1594 00:58:11,520 --> 00:58:14,240 the the Effect team has been great and a 1595 00:58:14,240 --> 00:58:16,320 lot of the projects I see written in it 1596 00:58:16,320 --> 00:58:17,780 are really cool. I think a lot of a lot 1597 00:58:17,780 --> 00:58:19,760 of people I respect are writing it. 1598 00:58:19,760 --> 00:58:21,120 And that's sort of 1599 00:58:21,120 --> 00:58:23,520 reason enough to try it. 1600 00:58:23,520 --> 00:58:25,840 I couldn't agree more. Thank you so much 1601 00:58:25,840 --> 00:58:28,240 for sharing your own experience with 1602 00:58:28,240 --> 00:58:30,800 Effect today and sharing your 1603 00:58:30,800 --> 00:58:32,560 super interesting story of like how 1604 00:58:32,560 --> 00:58:35,520 you've come to to co-found Warp. I 1605 00:58:35,520 --> 00:58:37,200 think it's very inspiring for the 1606 00:58:37,200 --> 00:58:39,840 audience to to show like how with how a 1607 00:58:39,840 --> 00:58:41,360 few engineers you 1608 00:58:41,360 --> 00:58:44,400 can build like a global system that that 1609 00:58:44,400 --> 00:58:47,136 solves like really material problems that 1610 00:58:47,136 --> 00:58:48,800 that's that's exactly what Effect is 1611 00:58:48,800 --> 00:58:52,107 built for and it's very cool to see that 1612 00:58:52,107 --> 00:58:53,280 is working well for you. 1613 00:58:53,280 --> 00:58:55,280 Yeah, we're we're super happy with 1614 00:58:55,280 --> 00:58:56,400 how it's turning out. 1615 00:58:56,400 --> 00:58:59,200 And if if anyone listening wants the 1616 00:58:59,200 --> 00:59:01,920 write Effect, come come join Warp. 1617 00:59:01,920 --> 00:59:03,920 Please do. Yeah, we're going to put the 1618 00:59:03,920 --> 00:59:06,560 link in the in the show notes and yeah, 1619 00:59:06,560 --> 00:59:08,880 Adam, thank you so much for taking the 1620 00:59:08,880 --> 00:59:10,560 time today and sharing your story. 1621 00:59:10,560 --> 00:59:12,240 Yeah, it was great to join. Thank you so 1622 00:59:12,240 --> 00:59:13,200 much for having me. Hopefully 1623 00:59:13,200 --> 00:59:14,640 we can do it again sometime. 1624 00:59:14,640 --> 00:59:16,640 That sounds great. All right. 1625 00:59:16,640 --> 00:59:17,920 Thank you for listening to the 1626 00:59:17,920 --> 00:59:19,520 Cause & Effect Podcast. 1627 00:59:19,520 --> 00:59:21,520 If you've enjoyed this episode, please 1628 00:59:21,520 --> 00:59:22,960 subscribe, leave a review 1629 00:59:22,960 --> 00:59:24,960 and share it with your friends. 1630 00:59:24,960 --> 00:59:26,800 If you haven't done so already, you can 1631 00:59:26,800 --> 00:59:28,400 join our Discord community. 1632 00:59:28,400 --> 00:59:30,320 And if you have any questions, feedback 1633 00:59:30,320 --> 00:59:32,800 or suggestions about this episode or 1634 00:59:32,800 --> 00:59:34,320 about Effect in general, 1635 00:59:34,320 --> 00:59:36,480 don't hesitate to get in touch. 1636 00:59:36,480 --> 00:59:50,480 See you in the next episode.