1 00:00:00,060 --> 00:00:00,660 Nikolay: Hello, hello. 2 00:00:00,660 --> 00:00:03,580 This is Postgres FM, episode number 105. 3 00:00:04,860 --> 00:00:05,940 My name is Nikolay. 4 00:00:06,180 --> 00:00:07,540 I'm the founder of Postgres.AI. 5 00:00:07,900 --> 00:00:11,620 And as usual, my co-host is Michael, the founder of pgMustard. 6 00:00:11,840 --> 00:00:12,600 Hi, Michael. 7 00:00:13,360 --> 00:00:14,200 Michael: Hello, Nikolay. 8 00:00:15,400 --> 00:00:17,460 Nikolay: So you chose an interesting topic. 9 00:00:18,060 --> 00:00:18,920 Is it interesting? 10 00:00:19,080 --> 00:00:19,800 I'm kidding. 11 00:00:20,320 --> 00:00:22,860 But it's not technical at all, right? 12 00:00:23,440 --> 00:00:25,440 Michael: It depends on how far we get into it. 13 00:00:25,640 --> 00:00:26,140 Nikolay: Yeah. 14 00:00:26,480 --> 00:00:31,960 So, we talk about startups, Postgres startups, ecosystem maybe 15 00:00:32,100 --> 00:00:36,720 for Postgres companies and new companies, young companies. 16 00:00:37,640 --> 00:00:40,680 We saw many of them launched in previous years. 17 00:00:41,940 --> 00:00:46,200 And recently, starting last year, we started observing some of 18 00:00:46,200 --> 00:00:48,920 them closing, which is natural for startups. 19 00:00:49,200 --> 00:00:50,580 Yeah, let's discuss this. 20 00:00:50,600 --> 00:00:51,640 Michael: Yeah, sounds great. 21 00:00:52,060 --> 00:00:53,520 I thought it might be interesting for people. 22 00:00:53,520 --> 00:00:57,160 I think not everybody pays as close attention to this as yourself 23 00:00:57,160 --> 00:00:57,880 and I do. 24 00:00:57,880 --> 00:01:00,040 So I think there'll be people out there that won't have heard 25 00:01:00,040 --> 00:01:01,400 of a bunch of these companies. 26 00:01:01,400 --> 00:01:03,840 I think it'd be really interesting to discuss what some of them 27 00:01:03,840 --> 00:01:07,480 are doing, what kind of categories there are, how they run, what 28 00:01:07,480 --> 00:01:10,200 kind of business models are popping up, what problems they're 29 00:01:10,200 --> 00:01:11,620 solving, that kind of thing. 30 00:01:11,720 --> 00:01:14,120 But the reason I brought this up now, I mean I'm kind of surprised 31 00:01:14,120 --> 00:01:16,800 we haven't done this kind of thing already, But the reason I 32 00:01:16,800 --> 00:01:20,060 brought this up now is I saw a tweet, maybe some other people, 33 00:01:20,060 --> 00:01:22,540 it got a few hundred likes, I think probably a few people in 34 00:01:22,540 --> 00:01:24,380 the Postgres world have seen it already. 35 00:01:25,600 --> 00:01:31,560 I think it was a semi-serious prediction, which was from a guy 36 00:01:31,560 --> 00:01:32,820 called Dax Rad. 37 00:01:33,160 --> 00:01:34,920 I don't know if I'm pronouncing it right. 38 00:01:34,920 --> 00:01:36,720 He's a bit of a joker online. 39 00:01:36,880 --> 00:01:39,420 It does quite a few funny tweets, but I think this was semi-serious. 40 00:01:39,920 --> 00:01:44,980 And it was a prediction that over the next 12 months, Postgres 41 00:01:45,060 --> 00:01:46,500 startups are going to get decimated. 42 00:01:46,840 --> 00:01:50,820 And that was the whole tweet, definitely in character for trolling. 43 00:01:51,280 --> 00:01:53,740 But I thought it was interesting and funny. 44 00:01:53,740 --> 00:01:56,720 And yeah, there were some funny responses too. 45 00:01:56,720 --> 00:02:00,360 But that prompted me to think, what is the Postgres startup ecosystem 46 00:02:00,480 --> 00:02:01,520 like right now? 47 00:02:01,560 --> 00:02:04,580 Nikolay: I think it's, first of all, it's serious, very serious. 48 00:02:05,900 --> 00:02:11,260 And second, I think it's wrong because it talks about next 12 49 00:02:11,400 --> 00:02:14,480 months, not talking about what already happened last 12 months. 50 00:02:14,480 --> 00:02:16,860 I think It's already happened. 51 00:02:18,400 --> 00:02:22,680 Because if you think like, even if 3 startups closed, that means 52 00:02:23,840 --> 00:02:25,520 they do have 30, right? 53 00:02:25,520 --> 00:02:28,520 Because what does this word mean? 54 00:02:29,240 --> 00:02:29,940 It means... 55 00:02:30,540 --> 00:02:31,400 Michael: I did look it up. 56 00:02:31,400 --> 00:02:34,460 I did look it up because that was part of the replies. 57 00:02:34,460 --> 00:02:36,800 But I think there are two definitions. 58 00:02:37,120 --> 00:02:42,440 One is the kind of the strict old school definition which was, 59 00:02:42,440 --> 00:02:45,720 you know, from the Roman times, 1 in ten… KB – Punishment. 60 00:02:45,720 --> 00:02:46,680 MH – …dies. 61 00:02:46,680 --> 00:02:47,360 Yeah, exactly. 62 00:02:47,540 --> 00:02:50,280 But I don't think that's what's generally meant when the word 63 00:02:50,280 --> 00:02:51,380 Nikolay: decimated is used in English. 64 00:02:51,380 --> 00:02:52,400 It's not just dice. 65 00:02:53,360 --> 00:02:55,580 Other 9 killed this guy. 66 00:02:57,260 --> 00:03:00,880 This is how it was originally, it was terrible, right? 67 00:03:01,780 --> 00:03:02,500 Michael: Well, it was like 68 00:03:02,500 --> 00:03:03,420 Nikolay: a group punishment. 69 00:03:04,340 --> 00:03:06,320 One Random guy is killed by others, right? 70 00:03:06,320 --> 00:03:10,680 So this, I don't know why we discuss it, it's not relevant maybe. 71 00:03:10,680 --> 00:03:13,400 Michael: It's very relevant to the tweet, right, decimated. 72 00:03:13,660 --> 00:03:16,720 But I think the more normal definition, the more normal use of 73 00:03:16,720 --> 00:03:21,180 the word decimated means like a whole, a lot of them will die. 74 00:03:21,180 --> 00:03:21,980 Nikolay: Reduction, right? 75 00:03:21,980 --> 00:03:23,100 Reduction of number of... 76 00:03:23,100 --> 00:03:23,800 Big reduction. 77 00:03:24,520 --> 00:03:30,600 Well, actually in the world of startups, usually 9 of 10 die, 78 00:03:30,740 --> 00:03:32,820 not 1 of 10, right? 79 00:03:33,100 --> 00:03:33,980 It's vice versa. 80 00:03:34,200 --> 00:03:38,300 Normally because lack of product market fit or something like 81 00:03:38,300 --> 00:03:43,860 that, bad execution and so on, which usually looks like we cannot 82 00:03:43,860 --> 00:03:45,980 raise next round, right? 83 00:03:46,720 --> 00:03:50,280 But back to my idea, what happened in the last 12 months. 84 00:03:50,280 --> 00:03:53,500 I know at least 3 startups which are dead now. 85 00:03:54,000 --> 00:03:56,760 Did we have 30 startups before that? 86 00:03:57,500 --> 00:04:00,540 If we did, I mean, did we have more? 87 00:04:00,860 --> 00:04:02,220 I think no, actually. 88 00:04:03,160 --> 00:04:04,540 Postgres-related startups. 89 00:04:04,660 --> 00:04:06,100 Michael: It depends how you define them. 90 00:04:06,100 --> 00:04:09,020 And I think even some of the ones you're talking about potentially 91 00:04:09,020 --> 00:04:10,900 talking about weren't Postgres only. 92 00:04:10,900 --> 00:04:11,980 So it depends... 93 00:04:11,980 --> 00:04:12,520 Nikolay: Of course. 94 00:04:12,520 --> 00:04:13,380 Nikolay: Of course. 95 00:04:13,380 --> 00:04:14,140 Nikolay: Of course. 96 00:04:14,140 --> 00:04:20,200 The biggest, aiven.com from Finland, I consider it a startup because 97 00:04:20,200 --> 00:04:25,080 I remember they were super small and then they skyrocketed and 98 00:04:25,080 --> 00:04:25,580 raised... 99 00:04:26,140 --> 00:04:28,860 They are Postgres related because they started, I think, from 100 00:04:28,860 --> 00:04:30,160 with Postgres and Kafka. 101 00:04:31,120 --> 00:04:36,180 It's like an alternative to AWS services, but multi-cloud like you. 102 00:04:36,280 --> 00:04:40,960 A single interface to multiple clouds, but including Postgres, right? 103 00:04:40,960 --> 00:04:41,020 right? 104 00:04:41,020 --> 00:04:45,060 And they had great Postgres materials, and so on. 105 00:04:45,060 --> 00:04:46,060 Michael: I think I misheard. 106 00:04:46,080 --> 00:04:47,520 What was the name of the company? 107 00:04:47,520 --> 00:04:48,620 How do you spell it? 108 00:04:48,840 --> 00:04:49,340 Nikolay: Aiven. 109 00:04:51,300 --> 00:04:51,800 Michael: Aiven. 110 00:04:52,720 --> 00:04:55,440 It must be a different Aiven to the one I'm thinking of because 111 00:04:55,440 --> 00:04:57,180 they seem to still be going strong. 112 00:04:58,140 --> 00:04:58,780 Nikolay: No, no, no, no, no. 113 00:04:58,780 --> 00:04:59,780 They didn't close. 114 00:05:00,040 --> 00:05:05,440 I mean, what I tried to say, I just reflected your words about 115 00:05:06,140 --> 00:05:12,620 not only are they strong, although I heard about layoffs, but 116 00:05:12,620 --> 00:05:17,140 they also raised the maximum number of VC capital among all, 117 00:05:17,840 --> 00:05:21,260 let's call it Postgres-related companies, right? 118 00:05:21,700 --> 00:05:25,120 So I just reflected the idea that some of them are not purely 119 00:05:25,120 --> 00:05:25,620 Postgres. 120 00:05:25,680 --> 00:05:27,720 And these guys are not purely Postgres. 121 00:05:27,720 --> 00:05:31,860 Of course, they have MySQL and a lot of elastic and everything, 122 00:05:31,860 --> 00:05:32,260 right? 123 00:05:32,260 --> 00:05:35,740 Yeah, I would say we have a big range of very big companies and 124 00:05:35,740 --> 00:05:41,540 Aiven, at least according to Crunchbase, raised $420 million. 125 00:05:43,780 --> 00:05:45,220 It's insane, right? 126 00:05:45,660 --> 00:05:46,860 Michael: With a lot of money. 127 00:05:46,920 --> 00:05:49,940 Nikolay: Yeah, I think they were first who reached the like kind 128 00:05:49,940 --> 00:05:54,440 of unicorn label among Postgres, let's call it Postgres-related 129 00:05:54,440 --> 00:05:55,340 startups. 130 00:05:56,000 --> 00:05:59,760 The other one I think is TimescaleDB also reached this good mark, 131 00:05:59,760 --> 00:06:00,020 right? 132 00:06:00,020 --> 00:06:04,580 Like unicorn means a valuation exceeding 1 billion. 133 00:06:05,320 --> 00:06:09,840 By the way, speaking of numbers, 420 is easy to remember sitting 134 00:06:09,840 --> 00:06:10,560 in California. 135 00:06:12,740 --> 00:06:16,680 But it's an impressive amount of money and we have Supabase 136 00:06:16,760 --> 00:06:19,740 and Timescale which raised more than 100 million, right? 137 00:06:20,460 --> 00:06:22,320 Michael: Neon too, over 100 million. 138 00:06:22,440 --> 00:06:23,910 Nikolay: Great, congrats everyone. 139 00:06:23,910 --> 00:06:25,020 Michael: Hasura as well. 140 00:06:25,680 --> 00:06:26,500 Nikolay: Yeah, Hasura. 141 00:06:26,600 --> 00:06:32,440 By the way, I stopped hearing about Hasura for a couple of years 142 00:06:32,440 --> 00:06:33,340 already somehow. 143 00:06:33,740 --> 00:06:36,340 And by the way, they are also starting from Postgres. 144 00:06:36,340 --> 00:06:39,180 I think they expanded to SQL Server and everything. 145 00:06:40,080 --> 00:06:46,560 Yeah, we didn't see the collapse of really big startups, but smaller 146 00:06:46,860 --> 00:06:52,160 and I think I would say Autotune managed to be already like mid-size 147 00:06:52,200 --> 00:06:54,440 because how much did they raise? 148 00:06:55,240 --> 00:06:56,100 16 million? 149 00:06:56,540 --> 00:06:57,040 15? 150 00:06:57,600 --> 00:07:01,900 Michael: I saw 12, but I don't know what they raised before that. 151 00:07:02,300 --> 00:07:05,140 Nikolay: I think 15 or 16 was the total I saw. 152 00:07:06,420 --> 00:07:10,680 Honestly, if we say database startups, this money should be not 153 00:07:10,680 --> 00:07:15,080 serious, because database startups require more money, I think, 154 00:07:15,660 --> 00:07:16,620 to go serious. 155 00:07:16,720 --> 00:07:19,180 And we have a lot of small, very small things. 156 00:07:19,340 --> 00:07:21,880 So let's talk about those who closed. 157 00:07:22,080 --> 00:07:24,560 Like Autotune is 1 of the latest news. 158 00:07:25,560 --> 00:07:26,300 Who else? 159 00:07:27,440 --> 00:07:30,840 Michael: There's only 1 other that I know of, and that was Snaplet. 160 00:07:31,420 --> 00:07:32,380 Nikolay: Snaplet, yeah. 161 00:07:32,380 --> 00:07:34,840 Michael: It's not really in the same category, but it's kind 162 00:07:34,840 --> 00:07:36,780 of in our space of making tools. 163 00:07:38,500 --> 00:07:39,640 Nikolay: What about Bit.io? 164 00:07:40,520 --> 00:07:42,380 Bit.io, it was a nice idea. 165 00:07:42,380 --> 00:07:46,880 You have a spreadsheet, you just throw it to browser uploading, 166 00:07:46,920 --> 00:07:47,360 right? 167 00:07:47,360 --> 00:07:51,700 And it immediately creates Postgres node with table and with 168 00:07:51,700 --> 00:07:51,880 data. 169 00:07:51,880 --> 00:07:52,560 Michael: Oh, really? 170 00:07:52,900 --> 00:07:53,160 Nikolay: Yeah. 171 00:07:53,160 --> 00:07:56,600 It's like super cloud approach, like very, very cloud. 172 00:07:56,600 --> 00:08:02,440 Like you just drop a spreadsheet and start running SQL queries 173 00:08:02,440 --> 00:08:03,120 to it. 174 00:08:03,260 --> 00:08:06,560 Was a nice idea, but they closed last year, I guess. 175 00:08:07,440 --> 00:08:12,340 And similarly to AutoTune, they did very, very, very, very, very, 176 00:08:12,340 --> 00:08:15,420 very short notice for their customers, which is, I think, very 177 00:08:15,420 --> 00:08:16,440 bad behavior. 178 00:08:16,860 --> 00:08:17,760 Not good, not good. 179 00:08:17,760 --> 00:08:19,540 Like 30 days or so. 180 00:08:20,140 --> 00:08:20,980 Get your data. 181 00:08:20,980 --> 00:08:21,480 Yes, 182 00:08:21,980 --> 00:08:23,260 Michael: it's difficult, isn't it? 183 00:08:23,260 --> 00:08:27,160 Nikolay: So these 3 startups, like Snaplet, Bit.io and Autotune 184 00:08:27,540 --> 00:08:29,120 are closed. 185 00:08:30,960 --> 00:08:33,880 And my question is how many startups do we have post-Gresley 186 00:08:34,020 --> 00:08:36,680 that are in this kind of broader term in general? 187 00:08:36,900 --> 00:08:38,100 Maybe 30, right? 188 00:08:38,160 --> 00:08:41,260 But 30 is quite a lot in my head. 189 00:08:41,840 --> 00:08:47,380 And it means that maybe we already have this mission complete, 190 00:08:47,520 --> 00:08:48,140 you know? 191 00:08:49,200 --> 00:08:49,700 Michael: So... 192 00:08:50,320 --> 00:08:52,940 Easiest prediction to make is something that's already happened. 193 00:08:53,480 --> 00:08:53,800 Nikolay: Right. 194 00:08:53,800 --> 00:08:57,280 Well, no, I mean, next 12 months maybe it will happen again. 195 00:08:58,440 --> 00:08:59,480 Michael: What do you think? 196 00:08:59,760 --> 00:09:00,560 Nikolay: I don't know. 197 00:09:00,780 --> 00:09:03,540 By the way, I somewhere had a list of startups. 198 00:09:05,660 --> 00:09:07,380 I think it didn't exceed 30. 199 00:09:07,900 --> 00:09:08,980 Michael: Yeah, I've made a list. 200 00:09:08,980 --> 00:09:11,760 I've made a bit of a list, but I didn't know what should count. 201 00:09:11,760 --> 00:09:14,760 Like, how many years does it still count as a startup? 202 00:09:15,060 --> 00:09:15,480 Nikolay: Right. 203 00:09:15,480 --> 00:09:16,500 What is startup? 204 00:09:17,020 --> 00:09:17,260 Michael: Yeah. 205 00:09:17,260 --> 00:09:19,600 What about companies that haven't gone down the VC route? 206 00:09:19,600 --> 00:09:21,420 That kind of like, do they count? 207 00:09:22,180 --> 00:09:24,100 So it's, I think it's really difficult. 208 00:09:24,140 --> 00:09:27,640 And then especially if you start to include companies that do. 209 00:09:28,580 --> 00:09:30,640 Tools for more than 1 type of database. 210 00:09:31,640 --> 00:09:35,500 Like We had an acquisition recently in the Postgres space where 211 00:09:35,500 --> 00:09:37,660 Timescale acquired a company called Popsicle. 212 00:09:38,360 --> 00:09:39,720 A cool name, by the way. 213 00:09:42,400 --> 00:09:45,920 But they did a tool that worked with Postgres, but not only Postgres. 214 00:09:45,920 --> 00:09:48,400 And there's been loads of companies like that, that I wouldn't 215 00:09:48,400 --> 00:09:50,780 have thought to include if I was making this list. 216 00:09:51,140 --> 00:09:54,260 So it's a very, very tricky definition. 217 00:09:54,520 --> 00:09:58,080 But to be fair, DAX did actually qualify a little bit in some 218 00:09:58,080 --> 00:10:01,720 replies, saying it was about the crop of companies that have 219 00:10:01,720 --> 00:10:05,500 popped up to offer managed Postgres services, which does limit 220 00:10:05,500 --> 00:10:05,660 it. 221 00:10:05,660 --> 00:10:06,840 Nikolay: It's very different. 222 00:10:07,360 --> 00:10:10,380 Michael: There can't be that many of those, maybe 5 or 10. 223 00:10:10,520 --> 00:10:11,020 Nikolay: Right. 224 00:10:11,880 --> 00:10:16,920 And yeah, we cannot take existing companies which started doing 225 00:10:16,920 --> 00:10:20,340 this because they already cannot be considered startups. 226 00:10:22,120 --> 00:10:26,140 So it means we talk about only like very few companies. 227 00:10:26,380 --> 00:10:31,200 I think not more than 10 because most managed Postgres services 228 00:10:31,200 --> 00:10:37,400 I can think of, they are done by companies which exist for many years. 229 00:10:37,400 --> 00:10:40,740 So, Supabase, Neon, Tembo, who else? 230 00:10:41,580 --> 00:10:42,700 Michael: Yeah, for sure. 231 00:10:42,700 --> 00:10:45,120 I don't know how some of the others are monetizing. 232 00:10:45,420 --> 00:10:50,140 Like, for example, there are people like FerretDB offering Mongo-compatible 233 00:10:50,440 --> 00:10:50,940 Postgres. 234 00:10:50,940 --> 00:10:53,540 I don't think they're offering a managed service though. 235 00:10:56,120 --> 00:10:57,700 Is PGEdge as well? 236 00:10:57,700 --> 00:10:59,160 Are they doing managed? 237 00:11:00,060 --> 00:11:01,180 Nikolay: I don't follow. 238 00:11:01,320 --> 00:11:02,320 Michael: I think so. 239 00:11:03,100 --> 00:11:06,140 PeerDB as well, I don't know if they're doing managed. 240 00:11:07,100 --> 00:11:09,020 I don't know if you would count it. 241 00:11:09,020 --> 00:11:12,120 Nikolay: They don't have managed Postgres, they have managed 242 00:11:12,120 --> 00:11:13,600 replication for Postgres. 243 00:11:13,940 --> 00:11:14,720 Michael: There you go. 244 00:11:15,400 --> 00:11:17,940 So I don't think they count in this prediction. 245 00:11:18,540 --> 00:11:20,640 So it's a tricky one to... 246 00:11:21,160 --> 00:11:25,600 I suspect it did its job and it was pure engagement farming and 247 00:11:25,600 --> 00:11:27,660 I've fallen for it. 248 00:11:27,660 --> 00:11:31,060 But in the grand scheme of things, I was most keen just to discuss 249 00:11:31,080 --> 00:11:33,520 how do you see Postgres startups at the moment? 250 00:11:33,520 --> 00:11:36,900 And are there any gaps that you'd really like to see maybe new 251 00:11:36,900 --> 00:11:40,320 startups pop up or what kind of problems some of these are solving? 252 00:11:40,320 --> 00:11:42,100 Which ones are most important to you? 253 00:11:42,100 --> 00:11:43,300 Like that kind of thing? 254 00:11:44,060 --> 00:11:47,360 Nikolay: Well I think I see things very differently than a few 255 00:11:47,360 --> 00:11:48,180 years ago. 256 00:11:49,300 --> 00:11:53,920 I think a lot of new people came, which is great, from different 257 00:11:54,480 --> 00:11:58,660 market segments, like even from database, from different database 258 00:11:58,660 --> 00:11:59,960 systems, for example, right? 259 00:11:59,960 --> 00:12:03,140 Like Tembo is 1 of great examples here. 260 00:12:03,700 --> 00:12:05,380 Michael: And, by the way, congrats to them. 261 00:12:05,380 --> 00:12:07,160 They just announced a raise. 262 00:12:08,040 --> 00:12:08,230 Nikolay: 14 million 263 00:12:08,230 --> 00:12:08,420 Michael: or something. 264 00:12:08,420 --> 00:12:11,560 Nikolay: And special congrats for Cybertruck to Ray. 265 00:12:14,020 --> 00:12:15,260 Michael: We'll link that up. 266 00:12:15,320 --> 00:12:18,280 Nikolay: With license plate startup, very relevant to today's 267 00:12:18,280 --> 00:12:19,260 discussion, right? 268 00:12:20,080 --> 00:12:25,760 I think we have very different types of companies and many attempts 269 00:12:25,760 --> 00:12:30,600 and it's interesting time right now because if we talk about 270 00:12:30,600 --> 00:12:34,600 VC maybe it's harder to raise but at the same time we know it's 271 00:12:34,600 --> 00:12:39,960 harder to find a job, which also means that it's easier to find 272 00:12:39,960 --> 00:12:40,460 people. 273 00:12:41,500 --> 00:12:45,300 And for those who make startups, it's some additional window 274 00:12:45,300 --> 00:12:45,940 of opportunity. 275 00:12:47,260 --> 00:12:49,860 So it's a mixed situation right now. 276 00:12:50,380 --> 00:12:54,940 Like you might fail to raise, but you might find good co-founder, 277 00:12:55,040 --> 00:12:56,880 for example, technical, right? 278 00:12:57,180 --> 00:12:59,520 And, and, and build something. 279 00:12:59,760 --> 00:13:00,040 Yeah. 280 00:13:00,040 --> 00:13:04,060 I'm honestly not standard startup guy at all. 281 00:13:04,780 --> 00:13:06,240 I'm slow, bootstrapping. 282 00:13:07,060 --> 00:13:09,440 I'm just observing this like VC. 283 00:13:09,780 --> 00:13:15,280 If you raise, for example, 10, 20 million dollars, build something 284 00:13:15,280 --> 00:13:20,080 which brings you some significant money, and then you need to 285 00:13:20,080 --> 00:13:25,600 shut down your company because you didn't have 10x growth per 286 00:13:25,600 --> 00:13:26,100 year. 287 00:13:27,040 --> 00:13:27,720 It's weird. 288 00:13:27,720 --> 00:13:28,480 It's weird. 289 00:13:29,340 --> 00:13:32,220 For VCs, it's probably like a failure, right? 290 00:13:33,740 --> 00:13:35,660 Because they expect hyper growth. 291 00:13:36,760 --> 00:13:41,100 But why do we need the hyper growth in post-growth ecosystem? 292 00:13:41,520 --> 00:13:45,060 Maybe we should have good companies which have normal growth, 293 00:13:45,060 --> 00:13:49,180 like 30% per year and that's it. 294 00:13:49,300 --> 00:13:51,320 I don't know, what do you think? 295 00:13:51,340 --> 00:13:54,000 Michael: Well, I'm in the same boat as you in terms of my own 296 00:13:54,000 --> 00:13:58,880 companies, trying to grow it sustainably with customer revenue, 297 00:13:59,620 --> 00:14:02,540 customer funded, bootstrapped at the beginning. 298 00:14:03,000 --> 00:14:06,860 So trying to do things slow and steady, but to do that, I felt 299 00:14:06,860 --> 00:14:10,800 like I had to pick a very, very small niche of a product, a very, 300 00:14:10,800 --> 00:14:14,480 very kind of low surface area compete in like a very, very small 301 00:14:14,480 --> 00:14:14,980 area. 302 00:14:15,360 --> 00:14:20,520 But I do think there are problems in databases that do require 303 00:14:20,740 --> 00:14:22,820 a lot more investment up front. 304 00:14:22,820 --> 00:14:26,540 And those things are harder to do self-funded slowly. 305 00:14:26,980 --> 00:14:29,140 I think they do require a lot of capital. 306 00:14:30,080 --> 00:14:34,600 And then the only model at the moment for that is either, well, 307 00:14:34,600 --> 00:14:38,940 I guess there are examples, even in the Postgres ecosystem of 308 00:14:38,940 --> 00:14:43,320 companies funding their own development there, like it, well, 309 00:14:43,320 --> 00:14:47,600 back to the hyperscalers, but even, So obviously companies like 310 00:14:47,600 --> 00:14:54,640 AWS, RDS, that was all funded through Amazon profits and Microsoft 311 00:14:54,720 --> 00:14:59,020 as well, Google, but in the Postgres space, Crunchy Data, for 312 00:14:59,020 --> 00:15:02,620 example, have been investing in their own platform that presumably 313 00:15:03,420 --> 00:15:08,160 largely funded through their own consulting business and then 314 00:15:08,160 --> 00:15:09,080 customer revenues. 315 00:15:09,300 --> 00:15:11,820 But it's like yours, right? 316 00:15:11,960 --> 00:15:14,540 Your business is also mixed consulting revenue. 317 00:15:14,540 --> 00:15:18,780 So some of these other companies I see are offering kind of service-based 318 00:15:19,160 --> 00:15:24,100 business models as well as the kind of more VC-friendly subscriptions 319 00:15:24,620 --> 00:15:27,680 or hosting, like usage fees. 320 00:15:27,880 --> 00:15:31,580 Nikolay: What is good about this all, it's definitely ecosystem 321 00:15:31,640 --> 00:15:36,640 is still growing, expanding, and we see the growth of companies 322 00:15:36,660 --> 00:15:40,340 which use Postgres in so many market segments. 323 00:15:41,320 --> 00:15:42,540 Everywhere, it's everywhere. 324 00:15:42,740 --> 00:15:43,500 It's great. 325 00:15:44,180 --> 00:15:49,140 We see many customers who need help in many aspects. 326 00:15:50,380 --> 00:15:56,400 So I agree we must divide companies by what they do, goals. 327 00:15:56,540 --> 00:16:01,080 If you want to change something in the database itself, like build 328 00:16:01,080 --> 00:16:07,720 a new type of Postgres itself or rewrite part of Postgres, it requires 329 00:16:07,720 --> 00:16:14,560 a lot of effort and you must have a very long vision. 330 00:16:15,040 --> 00:16:19,380 If you raise just $1,000,000, for example, and next year VC will 331 00:16:19,660 --> 00:16:23,740 say either raise or close basically because you don't have customers. 332 00:16:23,740 --> 00:16:25,380 This is not how it will work. 333 00:16:27,940 --> 00:16:33,700 Fundamental things require a lot of effort and maybe money also. 334 00:16:34,540 --> 00:16:35,920 So Neon is a good example. 335 00:16:35,920 --> 00:16:37,580 They change. 336 00:16:38,040 --> 00:16:39,040 Michael: Yeah, great example. 337 00:16:39,520 --> 00:16:43,500 Nikolay: Yeah, and also AlloyDB, which is still under development 338 00:16:43,580 --> 00:16:48,700 and Alexander Khodorkov joined Supabase also good driver probably 339 00:16:48,720 --> 00:16:53,700 to continue the development of AureolaDB because what is there we 340 00:16:54,000 --> 00:16:56,340 as Postgres users do need, right? 341 00:16:57,080 --> 00:17:00,320 Features which are designed there. 342 00:17:01,100 --> 00:17:04,780 Also, I would like to mention here like historical things. 343 00:17:05,820 --> 00:17:09,620 If maybe you've watched this, Michael Stonebraker, who is like 344 00:17:09,680 --> 00:17:15,700 father of Postgres, receiving Turing Award 10 years ago or so, 345 00:17:15,800 --> 00:17:21,080 he had a great lecture about startups in the database, database startups, 346 00:17:21,560 --> 00:17:25,320 because he is like, not only father of Postgres, but he's the father 347 00:17:25,320 --> 00:17:31,420 of database startups as well, participating in doing this a lot. 348 00:17:31,820 --> 00:17:33,400 So it was a great lecture. 349 00:17:33,400 --> 00:17:34,600 I highly recommend it. 350 00:17:34,600 --> 00:17:39,280 I remember he was comparing fundamental database startups as 351 00:17:39,280 --> 00:17:44,000 riding a bicycle across America, like downhill, uphill. 352 00:17:44,920 --> 00:17:46,220 Sometimes it feels great. 353 00:17:46,220 --> 00:17:49,120 Sometimes it's like almost killing you. 354 00:17:49,820 --> 00:17:51,180 Very good lecture actually. 355 00:17:51,380 --> 00:17:52,080 I enjoyed. 356 00:17:52,860 --> 00:17:54,960 So this is a fundamental thing. 357 00:17:54,960 --> 00:17:58,180 But there are other types of Postgres related startups as well 358 00:17:58,180 --> 00:18:04,540 and I think good things should happen when we saw in Postgres 359 00:18:04,540 --> 00:18:08,980 area, we saw the first acquisition by Microsoft, it was Citus, right? 360 00:18:09,140 --> 00:18:09,620 Michael: Yeah. 361 00:18:09,620 --> 00:18:09,920 Which is 362 00:18:09,920 --> 00:18:12,180 Nikolay: quite also fundamental, I would say, right? 363 00:18:12,180 --> 00:18:16,080 It's not Postgres itself, but it's a significant layer of addition 364 00:18:16,080 --> 00:18:16,740 to Postgres. 365 00:18:17,460 --> 00:18:20,420 Michael: It's actually the one area that I would like to see more 366 00:18:20,640 --> 00:18:22,200 deep work done. 367 00:18:22,500 --> 00:18:25,620 I think we've got a lot of bases covered in the Postgres ecosystem, 368 00:18:26,460 --> 00:18:30,980 but there is this kind of elephant in the room with PlanetScale 369 00:18:31,160 --> 00:18:31,820 and Vitesse. 370 00:18:32,860 --> 00:18:38,140 We had great guests on recently and I've since seen somebody 371 00:18:38,140 --> 00:18:41,460 describe, I think it was actually Arka from Notion, he described 372 00:18:41,600 --> 00:18:45,280 what Sammy and the Figma team are doing as somewhat starting 373 00:18:45,280 --> 00:18:47,300 to build the Vitesse for Postgres. 374 00:18:48,680 --> 00:18:51,500 I hadn't clicked that that's what they were doing, but that is 375 00:18:51,500 --> 00:18:52,680 kind of what they're doing. 376 00:18:52,720 --> 00:18:58,040 So that feels like a gap at the moment that could be, that probably 377 00:18:58,040 --> 00:19:01,960 does require a lot of development and could be funded by a VC-backed 378 00:19:02,120 --> 00:19:02,620 startup. 379 00:19:02,780 --> 00:19:07,260 But that's the one area that I don't see currently being done. 380 00:19:07,700 --> 00:19:08,000 Nikolay: Right. 381 00:19:08,000 --> 00:19:12,740 But then there are several acquisitions, I think, for smaller 382 00:19:12,980 --> 00:19:13,480 startups. 383 00:19:14,240 --> 00:19:18,140 And we saw also a few companies which are Postgres-related in 384 00:19:18,140 --> 00:19:18,640 YC. 385 00:19:19,980 --> 00:19:20,310 Yeah. 386 00:19:20,310 --> 00:19:21,640 Like PeerDB, right? 387 00:19:21,740 --> 00:19:23,200 Michael: Is that the first time, right? 388 00:19:23,200 --> 00:19:26,780 PeerDB and then ParadeDB, are they the first Postgres related 389 00:19:26,780 --> 00:19:27,880 companies to get into YC? 390 00:19:27,880 --> 00:19:29,340 Nikolay: No, Supabase was there as well. 391 00:19:29,340 --> 00:19:31,520 Michael: Oh, great shout, yeah. 392 00:19:31,520 --> 00:19:32,020 Nikolay: Good point. 393 00:19:32,020 --> 00:19:32,520 Someone else was there. 394 00:19:32,520 --> 00:19:34,020 Michael: Sorry, Supabase, I forgot. 395 00:19:34,480 --> 00:19:35,500 Nikolay: Yeah, I forgot. 396 00:19:35,500 --> 00:19:36,460 Maybe someone else. 397 00:19:36,460 --> 00:19:39,280 So, But there are bootstrappers as well. 398 00:19:39,280 --> 00:19:40,780 Is PgAnalyze bootstrapper? 399 00:19:41,120 --> 00:19:43,500 Michael: Yeah, I had a list of bootstrappers that I wanted to 400 00:19:43,500 --> 00:19:44,540 give a shout out to. 401 00:19:44,540 --> 00:19:47,460 Yeah, PgAnalyze is a great example, but also DBeaver. 402 00:19:48,840 --> 00:19:51,340 These bootstrappers don't get as much, because they don't raise 403 00:19:51,340 --> 00:19:54,580 money often, well never, they don't get as much press. 404 00:19:55,240 --> 00:19:57,840 So yeah, the rest of the other ones I had to give a shout out 405 00:19:57,840 --> 00:20:01,820 to were Egger Apps, who make Postico, My favorite little Mac 406 00:20:01,820 --> 00:20:02,940 client for Postgres. 407 00:20:03,520 --> 00:20:06,300 Postgres Compare, which is run by an old friend and colleague 408 00:20:06,300 --> 00:20:08,180 of mine, I call Neil Anderson. 409 00:20:08,760 --> 00:20:12,160 And also there's another monitoring company, so PG Analyze, of 410 00:20:12,160 --> 00:20:13,860 course, but with Lucas and the team. 411 00:20:13,860 --> 00:20:16,420 Nikolay: PG Analyze, I see it very often lately. 412 00:20:17,040 --> 00:20:17,540 Michael: Nice. 413 00:20:17,920 --> 00:20:21,360 Nikolay: Do you consider Core Root as a Postgres startup? 414 00:20:22,120 --> 00:20:24,180 It's very strange to classify, right? 415 00:20:24,620 --> 00:20:28,400 Because they have good support for Postgres for self-managed 416 00:20:28,580 --> 00:20:33,300 installations, I think, because they do eBPF monitoring, right? 417 00:20:34,020 --> 00:20:35,040 I like the idea. 418 00:20:35,060 --> 00:20:42,100 If we have like many Postgres nodes, and not only Postgres, we 419 00:20:42,100 --> 00:20:46,780 don't need to install anything like additional extensions, but 420 00:20:46,780 --> 00:20:49,940 we do need to have access to like... 421 00:20:50,280 --> 00:20:50,780 Mason. 422 00:20:51,180 --> 00:20:52,060 Well, right, right. 423 00:20:52,060 --> 00:20:54,120 So it's a self-managed situation. 424 00:20:55,080 --> 00:20:59,380 And they just observe whole traffic and perform a lot of interesting 425 00:20:59,380 --> 00:21:01,120 stuff, including like flame graphs. 426 00:21:01,940 --> 00:21:05,520 Historically, you choose the period and understand what Postgres 427 00:21:05,560 --> 00:21:09,060 did in that period in terms of what CPU was doing. 428 00:21:09,960 --> 00:21:12,660 And also query analysis. 429 00:21:14,160 --> 00:21:16,280 Michael: I forgot them, but I think it's because they're not 430 00:21:16,280 --> 00:21:18,240 purely, they're not only Postgres, right? 431 00:21:18,240 --> 00:21:18,760 So it's a different- 432 00:21:18,760 --> 00:21:24,440 Nikolay: Not only Postgres, but the founder was creating, it's 433 00:21:24,440 --> 00:21:28,440 the same founder who created Okimeter and it had very good Postgres 434 00:21:28,440 --> 00:21:30,840 monitoring, 1 of the best I consider still. 435 00:21:31,240 --> 00:21:34,640 And you know, I had last year, I had the overview of Postgres 436 00:21:34,640 --> 00:21:36,140 monitoring systems at PgCon. 437 00:21:37,440 --> 00:21:41,120 People liked a lot my slide deck, so I know what I'm talking 438 00:21:41,120 --> 00:21:41,620 about. 439 00:21:42,280 --> 00:21:45,140 And Peter Zaitsev joined him recently as co-founder. 440 00:21:46,160 --> 00:21:47,280 Michael: Yeah, very cool. 441 00:21:47,460 --> 00:21:51,040 Nikolay: So it means serious thoughts about databases, of course, 442 00:21:51,040 --> 00:21:56,100 maybe not Postgres only, but MySQL, but definitely serious about 443 00:21:56,100 --> 00:21:56,600 databases. 444 00:21:57,440 --> 00:21:59,980 So we can consider, like, this is the problem. 445 00:22:00,320 --> 00:22:04,540 Fundamental, not fundamental, bootstrapped, VC backed, Postgres 446 0:22:4,56 --> 0:22:5,9 only, not Postgres only. 446 00:22:05,900 --> 00:22:08,940 So many dimensions for this classification. 447 00:22:09,840 --> 00:22:12,870 Michael: Well, and that's part of what I wanted to discuss here. 448 00:22:12,870 --> 00:22:15,960 It was, it does feel quite rich at the moment. 449 00:22:15,960 --> 00:22:19,400 It does feel like we have a lot of options and I feel like that's 450 00:22:19,400 --> 00:22:22,760 really positive and there's quite a lot of innovation going on, 451 00:22:22,860 --> 00:22:24,360 not only at the startups. 452 00:22:24,440 --> 00:22:26,640 It feels like there's quite a lot of innovation going on even 453 00:22:26,640 --> 00:22:30,600 at some of the larger, older companies working on Postgres things. 454 00:22:30,700 --> 00:22:34,440 Maybe I'm biased and rose-tinted glasses, but it feels really 455 00:22:34,440 --> 00:22:36,000 positive to me at the moment. 456 00:22:36,400 --> 00:22:41,020 Nikolay: I think the idea of that tweet was the company who raised 457 00:22:41,020 --> 00:22:43,320 the lot are in danger. 458 00:22:43,940 --> 00:22:44,440 Michael: Interesting. 459 00:22:45,180 --> 00:22:45,880 You think? 460 00:22:46,260 --> 00:22:51,000 Nikolay: Yeah, because they need to raise harder and if they 461 00:22:51,000 --> 00:22:56,680 are not dealing with AI, for example, VCs won't give them money, 462 00:22:56,680 --> 00:22:57,180 right? 463 00:22:58,260 --> 00:22:58,760 Maybe. 464 00:22:58,860 --> 00:23:04,060 It's a very silly thought, but it's actually in many cases true. 465 00:23:04,760 --> 00:23:07,440 Michael: Well it will be interesting to see if any of them do 466 00:23:07,440 --> 00:23:12,100 but it also feels like the ones that I'm aware of, they have 467 00:23:12,100 --> 00:23:16,660 such good business models like if the business model is working, 468 00:23:17,360 --> 00:23:22,420 getting companies to host their database on your system, the 469 00:23:22,420 --> 00:23:25,880 ones that do well will grow with you and pay you more each month, 470 00:23:26,660 --> 00:23:27,600 month on month. 471 00:23:28,320 --> 00:23:30,060 And it's very difficult to move off, right? 472 00:23:30,060 --> 00:23:31,660 It's expensive, it's risky. 473 00:23:31,760 --> 00:23:33,660 So the fundamentals of... 474 00:23:34,160 --> 00:23:39,520 Nikolay: IMAGINE you grow only 50% per year or just 100% per 475 00:23:39,520 --> 00:23:39,880 year. 476 00:23:39,880 --> 00:23:42,240 For VCs, it's not enough at some point. 477 00:23:42,440 --> 00:23:43,160 Michael: Good point. 478 00:23:43,740 --> 00:23:44,520 Nikolay: It's tough. 479 00:23:45,520 --> 00:23:48,340 Michael: But if we're talking about 100 million in the bank, 480 00:23:48,340 --> 00:23:51,840 even if you hire a lot of people, that's a serious amount of 481 00:23:51,840 --> 00:23:52,340 money. 482 00:23:52,660 --> 00:23:53,160 Nikolay: Right. 483 00:23:54,740 --> 00:23:57,860 But how many managed Postgres services do we have already? 484 00:23:58,660 --> 00:24:02,280 And RDS is still the king of the hill, right? 485 00:24:03,340 --> 00:24:05,400 So I don't know. 486 00:24:06,820 --> 00:24:11,260 And is it still expanding the user base in terms of managed Postgres? 487 00:24:11,720 --> 00:24:12,440 Michael: Good question. 488 00:24:12,560 --> 00:24:17,440 I did look, going back to the tweet, I go for what, if that's 489 00:24:17,440 --> 00:24:19,220 the prediction, why is it the prediction? 490 00:24:19,440 --> 00:24:24,140 I personally think the number of databases is growing really 491 00:24:24,140 --> 00:24:24,520 quickly. 492 00:24:24,520 --> 00:24:27,440 The number of startups seems to be exploding at the moment because 493 00:24:27,440 --> 00:24:28,680 of the AI boom. 494 00:24:28,980 --> 00:24:35,380 A lot of the AI use cases do need a vector database. 495 00:24:35,380 --> 00:24:39,920 Postgres, because of pgVector, is positioned really well for vector 496 00:24:39,920 --> 00:24:39,960 databases. 497 00:24:39,960 --> 00:24:41,020 Nikolay: Do you think so? 498 00:24:42,040 --> 00:24:43,020 Michael: Personally, yes. 499 00:24:43,040 --> 00:24:44,620 Do you not think so? 500 00:24:45,020 --> 00:24:47,940 Nikolay: I have doubts, because it's not in core. 501 00:24:48,200 --> 00:24:50,740 If you check Postgres documentation, you don't see vector at 502 00:24:50,740 --> 00:24:51,240 all. 503 00:24:51,400 --> 00:24:56,660 This is a huge chance to lose this battle, in my opinion. 504 00:24:57,700 --> 00:24:58,600 This is serious. 505 00:24:59,340 --> 00:25:03,680 It might look so silly, simple, and so on, but it's a super serious 506 00:25:03,820 --> 00:25:05,880 problem Postgres currently has. 507 00:25:06,180 --> 00:25:07,900 And pgVector and all efforts. 508 00:25:07,900 --> 00:25:12,460 It's good that all major managed services quickly added pgVector, 509 00:25:12,600 --> 00:25:14,880 but they usually lag in terms of version. 510 00:25:15,620 --> 00:25:21,100 It means that they lag to deliver the best performance, for example, 511 00:25:21,100 --> 00:25:23,540 which is quickly improving in the latest versions. 512 00:25:24,520 --> 00:25:25,520 And so on and so on. 513 00:25:25,520 --> 00:25:30,460 Of course, this is a battle that's happening, but it's a barrier 514 00:25:30,460 --> 00:25:33,920 in minds, and some people, some folks will not pass this barrier. 515 00:25:34,000 --> 00:25:38,300 I mean, if you're not lazy, you will quickly realize it's kind 516 00:25:38,300 --> 00:25:40,740 of a zoo, it's a bazaar. 517 00:25:40,960 --> 00:25:42,980 It's open source, that's why, like, it's a bazaar. 518 00:25:42,980 --> 00:25:46,900 So we have many options, you need to quickly understand the options. 519 00:25:47,780 --> 00:25:51,900 But I recently had this customer, they said, okay, we are on 520 00:25:51,900 --> 00:25:57,340 RDS, I think it was Aurora, and we have pgVector, blah, blah, 521 00:25:57,340 --> 00:25:57,780 blah. 522 00:25:57,780 --> 00:26:02,360 And I see they use a very old version of pgVector, much older than 523 00:26:02,360 --> 00:26:06,520 even already available on Aurora, but they didn't upgrade because 524 00:26:06,820 --> 00:26:10,180 to upgrade you need to switch to a different… I don't remember. 525 00:26:10,360 --> 00:26:12,180 Some problems were happening. 526 00:26:12,660 --> 00:26:14,660 This whole thing is so fragmented. 527 00:26:15,170 --> 00:26:19,040 I mean vector story is very fragmented right now. 528 00:26:19,180 --> 00:26:25,360 And the only path to make it strong is to put it inside Postgres. 529 00:26:25,600 --> 00:26:28,640 But we know, we discussed that there are problems with that. 530 00:26:28,900 --> 00:26:30,400 Michael: Well, I don't think that's... 531 00:26:31,000 --> 00:26:32,280 I disagree personally. 532 00:26:32,540 --> 00:26:37,020 I think if you're really a startup and you can't keep pgVector 533 00:26:37,020 --> 00:26:42,540 up to date, you have bigger problems than, you know, that feels 534 00:26:42,540 --> 00:26:44,840 to me such an easy thing for you to do. 535 00:26:44,840 --> 00:26:48,040 If what I'm saying is true, If there are lots of companies that 536 00:26:48,040 --> 00:26:51,000 want vector databases right now and are willing to go with Postgres, 537 00:26:51,220 --> 00:26:54,020 I personally think it being out of core is a benefit right now 538 00:26:54,020 --> 00:26:57,320 because it means it can get bumps in version really quickly. 539 00:26:57,360 --> 00:27:00,920 Like, we see, I know they're not technically major versions, 540 00:27:00,920 --> 00:27:05,920 but 0.5, 0.6, 0.7 have all come with quite big functionality 541 00:27:06,160 --> 00:27:06,660 improvements. 542 00:27:06,680 --> 00:27:09,160 And if it was in core, we wouldn't have seen those. 543 00:27:09,160 --> 00:27:10,400 It would have taken years. 544 00:27:10,400 --> 00:27:14,000 So personally, I think there are advantages in that area. 545 00:27:14,280 --> 00:27:16,880 I've just remembered, though, I actually think the prediction 546 00:27:16,880 --> 00:27:20,580 was made partly because of 2 other reasons. 547 00:27:20,600 --> 00:27:26,720 I think it's partly that PlanetScale is very marketing level 548 00:27:26,720 --> 00:27:31,120 kind of attractive to kind of medium-sized, like non-AI kind 549 00:27:31,120 --> 00:27:32,760 of SaaS businesses. 550 00:27:33,140 --> 00:27:36,020 I don't know how many non AI businesses are popping up right 551 00:27:36,020 --> 00:27:40,120 now, but the ones that are, I think that they offer a very attractive 552 00:27:40,240 --> 00:27:44,160 proposition, like kind of like Mongo did before, but with SQL. 553 00:27:44,600 --> 00:27:47,720 And then I think the other driving force for the prediction is 554 00:27:47,720 --> 00:27:58,140 probably the number of SQLite companies popping up saying, actually, 555 00:27:58,140 --> 00:27:59,520 all you need is this. 556 00:27:59,540 --> 00:28:02,800 For a long, long time, you can get by with SQLite, keep things 557 00:28:02,800 --> 00:28:03,300 simple. 558 00:28:03,740 --> 00:28:09,140 So I think at least that part of Twitter that I'm seeing is talking 559 00:28:09,140 --> 00:28:10,900 about those topics a bit more. 560 00:28:10,900 --> 00:28:14,120 And that might be why they think the Postgres startups might 561 00:28:14,120 --> 00:28:18,840 struggle rather than a more nuanced take on vector databases. 562 00:28:19,080 --> 00:28:21,140 But I might have totally misunderstood. 563 00:28:21,540 --> 00:28:22,480 What do you think? 564 00:28:22,920 --> 00:28:23,800 Nikolay: Hard to say. 565 00:28:24,800 --> 00:28:25,740 I don't know. 566 00:28:25,920 --> 00:28:28,400 Michael: I just remembered there's 1 I completely forgot. 567 00:28:28,520 --> 00:28:29,640 Have you heard of Nile? 568 00:28:29,840 --> 00:28:30,340 N-I-L-E. 569 00:28:31,560 --> 00:28:32,060 Yeah. 570 00:28:32,860 --> 00:28:36,260 Because they popped up quite, like relatively recently in the 571 00:28:36,260 --> 00:28:40,640 grand scheme of things, talking about running a regular SaaS 572 00:28:40,640 --> 00:28:43,480 on Postgres and some innovations in that area. 573 00:28:43,480 --> 00:28:45,920 So that could be interesting, but they're another 1 that we're 574 00:28:45,920 --> 00:28:47,700 offering a managed service again. 575 00:28:49,540 --> 00:28:54,480 Nikolay: Well, yeah, talking about this lack of being in core, 576 00:28:54,480 --> 00:28:58,200 of course, I agree with you about the pace of development. 577 00:28:58,320 --> 00:28:59,020 That's great. 578 00:28:59,100 --> 00:29:01,900 And maybe things will change in the future. 579 00:29:02,920 --> 00:29:10,320 But it's just so sad that everyone has everything in documentation, 580 00:29:10,640 --> 00:29:16,120 like Techmonger, Elastic, everyone, like any major database system. 581 00:29:16,980 --> 00:29:18,620 They talk about vectors. 582 00:29:18,740 --> 00:29:21,720 Postgres documentation doesn't talk about vectors at all, right? 583 00:29:22,000 --> 00:29:27,100 I mean, it talks about arrays, but it's not like different, right? 584 00:29:27,620 --> 00:29:30,280 So maybe at some point it will go to core. 585 00:29:30,860 --> 00:29:32,380 I hope it won't be late. 586 00:29:33,340 --> 00:29:33,820 Right? 587 00:29:33,820 --> 00:29:37,320 Michael: But, just to push you a little bit, the super-based 588 00:29:37,420 --> 00:29:40,013 documentation talks about vectors, the neon documentation. 589 00:29:40,013 --> 00:29:40,440 Of course. 590 00:29:40,440 --> 00:29:42,440 Nikolay: RDS, Cloud SQL, all of them. 591 00:29:42,440 --> 00:29:42,940 Michael: Yeah. 592 00:29:43,320 --> 00:29:45,360 But people are still then going with Postgres. 593 00:29:45,480 --> 00:29:46,860 So what's the problem? 594 00:29:47,440 --> 00:29:51,740 Nikolay: The problem is that you need to spend time to choose 595 00:29:51,740 --> 00:29:55,860 and understand all the options here, because options are quite 596 00:29:56,840 --> 00:29:58,640 different everywhere. 597 00:29:59,480 --> 00:30:01,820 Michael: But that would be true even if it was in the Postgres 598 00:30:01,920 --> 00:30:02,860 docs too. 599 00:30:03,080 --> 00:30:04,420 In fact, 1 more option. 600 00:30:05,020 --> 00:30:06,040 Nikolay: Maybe, maybe. 601 00:30:07,660 --> 00:30:09,240 Michael: Anything you would like to see? 602 00:30:09,240 --> 00:30:12,160 Anything, like if anybody's listening, wanting to do a startup 603 00:30:12,160 --> 00:30:15,240 or like get inspired, is there any kind of area you'd like to 604 00:30:15,240 --> 00:30:16,600 see more innovation in? 605 00:30:16,740 --> 00:30:20,320 Nikolay: Of course, And by the way, Tembo is doing good steps 606 00:30:20,320 --> 00:30:21,180 in this direction. 607 00:30:21,760 --> 00:30:31,420 So I think, you know, 37signals idea about moving out of cloud. 608 00:30:33,420 --> 00:30:33,580 I 609 00:30:33,580 --> 00:30:35,320 Michael: thought you were going to say, I thought you were going 610 00:30:35,320 --> 00:30:36,240 to say moving. 611 00:30:36,340 --> 00:30:40,020 So they've also done this once product that runs on SQLite, like, 612 00:30:40,080 --> 00:30:42,520 but no, yeah, the moving out of the cloud 1. 613 00:30:42,520 --> 00:30:42,740 Yes. 614 00:30:42,740 --> 00:30:42,940 Nikol gaming 615 00:30:42,940 --> 00:30:47,860 Moving out of cloud and having greater level of control over 616 00:30:47,860 --> 00:30:48,720 your database. 617 00:30:50,140 --> 00:30:53,160 Direct access to files, for example, and so on, which is not 618 00:30:53,160 --> 00:30:55,440 possible for managed Postgres services usually. 619 00:30:55,680 --> 00:30:57,100 We discussed it a few times. 620 00:30:57,100 --> 00:30:59,980 Good thing that, for example, CrunchyBridge provides full-fledged 621 00:31:00,480 --> 00:31:01,320 super user. 622 00:31:02,860 --> 00:31:10,600 So I expect, actually, Cytos co-founders, they launched new startup 623 00:31:10,600 --> 00:31:11,480 as well, right? 624 00:31:11,480 --> 00:31:13,220 And they do Postgres as well. 625 00:31:13,680 --> 00:31:14,820 Yeah, good point. 626 00:31:14,820 --> 00:31:16,120 In UbiCloud, right? 627 00:31:16,120 --> 00:31:17,140 Michael: That was it, yeah. 628 00:31:17,780 --> 00:31:19,520 Forgot them, yeah, good point. 629 00:31:19,740 --> 00:31:20,500 Nikolay: The idea is great. 630 00:31:20,500 --> 00:31:23,600 You can install it anywhere, you have full access, but it's still, 631 00:31:23,800 --> 00:31:29,220 it's like the simplicity of management is similar to you have 632 00:31:29,220 --> 00:31:29,880 in cloud. 633 00:31:30,040 --> 00:31:33,220 You can do it with Superbase as well, by the way, I think. 634 00:31:33,260 --> 00:31:35,340 You can host yourself, it's great. 635 00:31:36,020 --> 00:31:42,040 So in this area, I expect good shifts in the future, because 636 00:31:42,040 --> 00:31:43,680 I think demand will be growing. 637 00:31:43,680 --> 00:31:50,780 People who are like, already say like, we've been in cloud enough. 638 00:31:50,920 --> 00:31:55,280 We want to reduce overhead, how much we spent and so on. 639 00:31:55,560 --> 00:31:59,720 Of course, some people will say cloud is great, I understand 640 00:31:59,720 --> 00:32:04,540 this, But I think this party will grow as well, like an anti-cloud 641 00:32:04,600 --> 00:32:06,140 party, let's say. 642 00:32:06,680 --> 00:32:09,080 Michael: I think it might also be a little bit like stage of 643 00:32:09,080 --> 00:32:09,400 company. 644 00:32:09,400 --> 00:32:12,580 I can imagine a lot of startups going with the cloud initially 645 00:32:12,660 --> 00:32:16,280 and then like that, perhaps a bit earlier than 37signals did, 646 00:32:16,280 --> 00:32:18,820 maybe then moves workloads back off. 647 00:32:18,820 --> 00:32:21,180 So I could definitely see both being true. 648 00:32:22,100 --> 00:32:22,600 Nikolay: Yes. 649 00:32:22,720 --> 00:32:26,120 So for example, new and fully betting, everything should be cloud. 650 00:32:26,120 --> 00:32:29,440 I understand that, but I also understand like then we pay like 651 00:32:29,440 --> 00:32:31,100 5 times more than we could. 652 00:32:31,920 --> 00:32:33,900 So why do we do this? 653 00:32:34,760 --> 00:32:38,800 I know small companies which also understand that they don't 654 00:32:38,800 --> 00:32:43,040 want to pay a lot and they are good technically, so they are 655 00:32:43,180 --> 00:32:45,780 successful and happy to be not in cloud. 656 00:32:46,000 --> 00:32:48,540 Or in very low-cost cloud, for example. 657 00:32:48,760 --> 00:32:52,280 And Postgres management there is interesting because you want 658 00:32:52,280 --> 00:32:54,880 all the features but don't want all the overhead. 659 00:32:55,580 --> 00:32:56,440 This is interesting. 660 00:32:56,740 --> 00:33:00,040 In general, I don't know what will happen, but it's normal for 661 00:33:00,040 --> 00:33:03,940 startups to die, especially if they are VC backed. 662 00:33:04,920 --> 00:33:10,580 Yeah, because VCs put a lot of pressure, especially after like 663 00:33:10,580 --> 00:33:13,300 3 to 5 years of existence of the company. 664 00:33:14,160 --> 00:33:19,500 I learned it hard way with my second startup, like 10 years ago, 665 00:33:20,020 --> 00:33:24,240 when I realized like 3 years passed and we need to move on and 666 00:33:24,520 --> 00:33:27,100 huge optimization was needed. 667 00:33:27,380 --> 00:33:29,940 It was successfully done, but it was very painful. 668 00:33:30,660 --> 00:33:36,420 So it's good if you know this in advance, that there is some 669 00:33:37,660 --> 00:33:41,920 period of time when you will need to raise again or you will 670 00:33:41,920 --> 00:33:42,840 just be closed. 671 00:33:44,040 --> 00:33:47,740 In my case it was not closure, it was just a huge optimization 672 00:33:47,840 --> 00:33:50,140 to be very profitable at some point. 673 00:33:50,740 --> 00:33:53,860 That's why I like the bootstrapped cases as well. 674 00:33:54,140 --> 00:33:54,640 Michael: Yeah. 675 00:33:54,640 --> 00:33:55,140 Right. 676 00:33:55,460 --> 00:33:57,720 And I'm glad we got a chance to shout some of them out because 677 00:33:57,720 --> 00:33:59,700 they don't always get the attention they deserve. 678 00:33:59,800 --> 00:34:02,760 Nikolay: If bootstrapped company doesn't die in first year and 679 00:34:02,760 --> 00:34:06,420 the founders still believe in this idea, it's a much more stable 680 00:34:06,420 --> 00:34:09,740 company actually, because they don't have pressure from VCs. 681 00:34:10,240 --> 00:34:14,160 But probably they have much less chances to grow significantly, 682 00:34:14,540 --> 00:34:15,040 right? 683 00:34:16,600 --> 00:34:18,020 Michael: Well, it depends, right? 684 00:34:18,080 --> 00:34:22,680 I actually forgot a couple of similar companies that are bootstrapped 685 00:34:22,680 --> 00:34:25,440 or mostly bootstrapped or at least they've stopped raising and 686 00:34:25,440 --> 00:34:26,660 they're doing really well. 687 00:34:27,440 --> 00:34:30,980 Like JetBrains, a really, really old company going from strength 688 00:34:30,980 --> 00:34:31,660 to strength. 689 00:34:31,780 --> 00:34:33,880 Nikolay: King of bootstrapped companies, I think. 690 00:34:35,200 --> 00:34:37,780 Michael: There are some examples of companies that have scaled 691 00:34:38,100 --> 00:34:43,360 to the same levels as some VC-backed companies without needing 692 00:34:43,360 --> 00:34:43,860 to. 693 00:34:44,540 --> 00:34:46,520 So kudos to them as well. 694 00:34:47,020 --> 00:34:47,520 Nikolay: Right. 695 00:34:47,540 --> 00:34:52,900 So my conclusion is like, I expect more closures and it's interesting 696 00:34:52,900 --> 00:34:55,160 to observe the situation. 697 00:34:55,640 --> 00:34:58,740 And unfortunately, I think closures will happen with companies 698 00:34:58,740 --> 00:35:04,900 who raised a lot Or they will need to change things a lot. 699 00:35:05,540 --> 00:35:06,040 Michael: Interesting. 700 00:35:06,100 --> 00:35:08,540 Well, I'm not, I would be surprised. 701 00:35:08,600 --> 00:35:12,320 So call me naive, but my prediction would be that none of them 702 00:35:12,320 --> 00:35:14,520 do close at least in the next 12 months. 703 00:35:14,960 --> 00:35:16,400 So let's see. 704 00:35:16,500 --> 00:35:18,040 I'll be rooting for them anyway. 705 00:35:18,280 --> 00:35:21,600 In fact, all the Postgres startups, I guess we'll be rooting 706 00:35:21,600 --> 00:35:22,100 for. 707 00:35:22,440 --> 00:35:26,080 Nikolay: Like a system is quite interesting and with challenges, 708 00:35:26,180 --> 00:35:27,840 but it's still growing, I think. 709 00:35:27,840 --> 00:35:31,400 And I think it's interesting to observe the further growth. 710 00:35:31,900 --> 00:35:34,220 Michael: Yeah, should be fun anyway. 711 00:35:34,900 --> 00:35:36,020 Thanks so much, Nikolay. 712 00:35:36,220 --> 00:35:37,540 Anything else you wanted to add? 713 00:35:37,540 --> 00:35:38,400 Nikolay: No, thank you. 714 00:35:38,400 --> 00:35:39,440 See you next time. 715 00:35:39,720 --> 00:35:40,840 Michael: See you next time.