1 00:00:00,901 --> 00:00:03,817 How do you make your data queryable using natural language? 2 00:00:03,817 --> 00:00:12,174 We see lots of terms pop up, RAG systems, semantic querying, and we'll show you today with MindsDB how to set that up in a few minutes. 3 00:00:12,174 --> 00:00:13,019 Let's go. 4 00:00:21,018 --> 00:00:24,443 Hi everyone, welcome to technology explorations at Dataminded. 5 00:00:24,443 --> 00:00:28,588 In this series, we give you an initial look in new or interesting technologies. 6 00:00:28,709 --> 00:00:34,397 My name is Jonny, Knowledge lead here at Dataminded, and today we'll have a look into MindsDB. 7 00:00:34,397 --> 00:00:36,260 And for that, I've invited Tarik. 8 00:00:36,260 --> 00:00:37,521 Welcome Tarik. 9 00:00:37,650 --> 00:00:39,334 Hi Jonny, thank you for having me. 10 00:00:39,334 --> 00:00:41,292 happy to be here to talk about MindsDB. 11 00:00:41,292 --> 00:00:50,460 So we will be building on top of what we did last time where we went over extracting data from Google Drive and putting it into a Postgres database and today you're going to show 12 00:00:50,460 --> 00:00:51,862 me MindsDB. 13 00:00:51,862 --> 00:00:53,622 Tell me what is MindsDB? 14 00:00:54,624 --> 00:01:04,788 uh MindsDB, we could call it an open source AI layer, which makes your data queryable using natural language. 15 00:01:04,849 --> 00:01:09,572 So what it can do is that it can connect to a database, like Postgres, for example. 16 00:01:09,572 --> 00:01:15,254 And then you can add AI capabilities using SQL commands. 17 00:01:15,873 --> 00:01:22,507 So what it can do is it turns your data into a knowledge base. 18 00:01:22,847 --> 00:01:31,482 And then you can build chatbots to talk questions about this MindsDB can also connect to multiple connectors. 19 00:01:31,739 --> 00:01:35,222 such as Slack, Notion, Databricks, and so on. 20 00:01:35,222 --> 00:01:37,393 So you can directly chat with your data. 21 00:01:37,393 --> 00:01:46,697 So you have two ways to use MindsDB You can use it to connect to a Postgres database that you own where you already ingested data, or you can directly connect to the data that 22 00:01:46,697 --> 00:01:49,359 lives in a system and query that. 23 00:01:49,902 --> 00:02:01,553 Okay, so MindsDB, if I understand correctly, is a layer that you use to fetch data from an existing system and then basically index it so that it's easily queryable by agents. 24 00:02:01,553 --> 00:02:02,963 Yes, Indeed. 25 00:02:02,963 --> 00:02:03,124 Is 26 00:02:03,124 --> 00:02:11,022 What we're going to do today is we are going to create a knowledge base on MindsDB from the database that we have in Postgres. 27 00:02:11,022 --> 00:02:15,013 a knowledge base is like a semantic search engine over the data. 28 00:02:15,013 --> 00:02:17,815 So it combines an embedding model, vector storage. 29 00:02:17,815 --> 00:02:24,389 metadata tracking And then it automatically chunks long documents for better retrieval. 30 00:02:24,489 --> 00:02:32,420 So it's a way for the chatbot or the LLM to have a better semantic understanding of the data that is behind it. 31 00:02:32,420 --> 00:02:36,328 Yeah, can you remind us again, what is the data we have available? 32 00:02:36,328 --> 00:02:47,834 so during the PyAirByte demo, we ingested a couple of text files from Google Drive, which now live in the Postgres table here in the Google Drive files. 33 00:02:47,834 --> 00:02:54,797 you have the names here of the files, and then you have the contents of the files here. 34 00:02:54,839 --> 00:03:05,128 So we will vectorize these, create a knowledge base out these files in the database, and then use a chatbot to ask questions about these files. 35 00:03:05,542 --> 00:03:06,795 Yeah, okay. 36 00:03:07,058 --> 00:03:08,953 Let's have a look at what it can do. 37 00:03:09,025 --> 00:03:10,436 this is mindsDB. 38 00:03:10,857 --> 00:03:18,402 Here, it's already running on my laptop, on this local host, locally. 39 00:03:18,402 --> 00:03:30,530 And then you can see here, for example, if you wanted to connect to a data source, you can have multiple data source in there where you can directly connect to the data, through the 40 00:03:30,530 --> 00:03:31,310 API. 41 00:03:32,493 --> 00:03:34,749 and then ask questions uh in real time. 42 00:03:34,749 --> 00:03:39,179 Or you can also create knowledge base from a selected data source. 43 00:03:39,179 --> 00:03:42,076 um 44 00:03:42,076 --> 00:03:50,267 a knowledge base, does that mean then reading all of this data and making embeddings out of it or chunking everything? 45 00:03:51,370 --> 00:03:52,000 Okay. 46 00:03:52,000 --> 00:04:01,189 and then you can select your own embedding models to do that and we will see how we can use the MindsDB to do it. 47 00:04:01,189 --> 00:04:04,572 You will notice that for example, there is no Google Drive connector. 48 00:04:04,572 --> 00:04:06,033 which in... 49 00:04:06,713 --> 00:04:07,198 Yes. 50 00:04:07,198 --> 00:04:08,539 to postgres. 51 00:04:08,623 --> 00:04:10,344 And I did see a Slack connector. 52 00:04:10,344 --> 00:04:16,639 So I see a mixture between data systems and also software systems or SaaS providers. 53 00:04:18,041 --> 00:04:18,410 Okay. 54 00:04:18,410 --> 00:04:20,734 I think there's also Teams, Notion. 55 00:04:20,734 --> 00:04:25,916 So we won't have time to explore everything, but there is a lot that you can do with it. 56 00:04:25,916 --> 00:04:30,838 You can also connect it to an MCP that talks to other data sources or providers. 57 00:04:30,838 --> 00:04:33,409 So it's a quite versatile tool. 58 00:04:34,036 --> 00:04:36,979 Okay, so you could have MCP as a source. 59 00:04:37,480 --> 00:04:39,442 And so MindsDB could read from there. 60 00:04:39,442 --> 00:04:42,685 We have another video on MCP for those who want to explore, 61 00:04:43,707 --> 00:04:46,008 We will go through the code later. 62 00:04:46,008 --> 00:04:49,791 But first, let's have a look at what is inside the MindsDB. 63 00:04:49,791 --> 00:04:55,734 Here we have, as you can see in my data sources, I already ingested the files. 64 00:04:56,394 --> 00:04:57,935 And they are there. 65 00:04:57,935 --> 00:04:59,475 Google Drive files. 66 00:05:00,209 --> 00:05:01,390 you can see the contents. 67 00:05:01,390 --> 00:05:06,113 This is basically what we see in Postgres. 68 00:05:06,113 --> 00:05:10,075 And then what we can do is that we create a knowledge base out of this. 69 00:05:10,075 --> 00:05:11,376 So it's here. 70 00:05:11,536 --> 00:05:20,000 And then here you have information about the name of the knowledge base, about the embedding models that we choose. 71 00:05:20,000 --> 00:05:21,931 Here is this is the one that I use. 72 00:05:21,931 --> 00:05:24,323 This is my API key, which is hidden. 73 00:05:24,323 --> 00:05:29,465 um And then you have metadata column and... 74 00:05:29,741 --> 00:05:30,962 parameters and so on. 75 00:05:30,962 --> 00:05:34,443 And what we do afterwards is that we have an agent here. 76 00:05:34,443 --> 00:05:36,684 So it's nicely structured here. 77 00:05:36,684 --> 00:05:40,947 You have your agents, you have your jobs, you have knowledge bases, models. 78 00:05:40,947 --> 00:05:50,092 You can also use local models if you so choose, which is quite powerful because if you want to do everything locally or you have sensitive files that you don't want to connect 79 00:05:50,092 --> 00:05:50,802 to. 80 00:05:50,943 --> 00:05:54,636 a cloud-based LLM, you can also use it locally. 81 00:05:54,636 --> 00:06:03,082 And then we have our agent here, which also uses a model and has built-in parameters like you need to connect to that knowledge base. 82 00:06:03,082 --> 00:06:04,682 You have a prompt template. 83 00:06:04,682 --> 00:06:06,023 You have a role. 84 00:06:06,023 --> 00:06:10,766 You answer user questions about that knowledge base, Google Drive knowledge base here. 85 00:06:10,766 --> 00:06:12,367 You should add the ID. 86 00:06:13,186 --> 00:06:21,550 Here when you quote or paraphrase a chunk, you should write the ID and the source, summarize duplicates and so on. 87 00:06:21,550 --> 00:06:32,556 So you're creating an agent with uh a specific prompting that connects to a knowledge base that you created, that is linked to files on the database so that you can use natural 88 00:06:32,556 --> 00:06:34,598 language to ask questions about your files. 89 00:06:34,598 --> 00:06:38,063 Yeah, yes. 90 00:06:38,063 --> 00:06:39,484 resides in the source. 91 00:06:39,484 --> 00:06:45,170 Then you have your knowledge base that is built up from data in the source and MindsDB helps you create that. 92 00:06:45,170 --> 00:06:54,035 And then you have an agent level, which can interact with that knowledge base, but also using an external agent to answer questions. 93 00:06:54,035 --> 00:06:55,027 Yes, exactly. 94 00:06:55,027 --> 00:07:00,082 So here I'm going to use the Google Drive agent to ask questions about my files. 95 00:07:00,082 --> 00:07:04,286 So here I will ask him to give me pros and cons about Microsoft Fabric, 96 00:07:04,286 --> 00:07:13,006 So here it's using the Google Drive Knowledge base because I've instructed him to do it. 97 00:07:14,206 --> 00:07:20,306 And he's doing SQL queries on the actual copy of the text. 98 00:07:20,306 --> 00:07:25,826 So here we have some pros and cons about Microsoft Fabric based on recent learnings. 99 00:07:25,826 --> 00:07:31,166 Here you see it's a fully managed platform and upgrade on the Power BI platform and so on. 100 00:07:31,166 --> 00:07:34,354 So you have a good view and then you have the source. 101 00:07:34,354 --> 00:07:40,402 the ID of the file and also the name of the file, which is the Facts and Breakfast, Celebrating Fabric's First Year. 102 00:07:40,402 --> 00:07:43,846 And we know that it's actually here, right? 103 00:07:43,846 --> 00:07:47,870 So it took the information from this file here. 104 00:07:48,311 --> 00:07:57,045 What I already wondered like is we see the query there select star from, and it's trying to do a where query with like a specific string match. 105 00:07:57,165 --> 00:08:05,409 This is translated into like an embedding behind the scenes that we don't see it, or is it actually trying to do like a literal match? 106 00:08:06,186 --> 00:08:08,208 No, he's doing semantic search. 107 00:08:09,270 --> 00:08:11,922 Yes, yes, yes, yes. 108 00:08:11,922 --> 00:08:13,375 Yeah, indeed. 109 00:08:13,375 --> 00:08:24,558 It's doing semantic search and then looking over all the vectors in the Google Drive knowledge base and then selecting the file which is the most appropriate to the question 110 00:08:24,558 --> 00:08:25,609 that we're asking. 111 00:08:25,814 --> 00:08:28,466 Okay, so the equal sign here is a bit misleading. 112 00:08:28,466 --> 00:08:32,930 It's actually doing a semantic search on the content column in this case. 113 00:08:34,111 --> 00:08:35,452 Okay, got it. 114 00:08:36,022 --> 00:08:43,912 So maybe I can show you the code on how to get this MindsDB up and running 115 00:08:43,914 --> 00:08:47,036 basically we're doing a Docker compose mindsDB. 116 00:08:47,036 --> 00:08:51,680 So we are getting a mindsDB image from mindsDB. 117 00:08:51,680 --> 00:08:56,601 What is really nice about that is that you don't have to log in anywhere. 118 00:08:56,601 --> 00:09:02,357 It's open source and you just download the image as a Docker, Docker compose up. 119 00:09:02,357 --> 00:09:05,318 and then you set up a username and a password and you're in. 120 00:09:05,318 --> 00:09:08,041 And you can use the full MindsDB experience. 121 00:09:08,041 --> 00:09:10,222 here I tried to code a little bit. 122 00:09:10,222 --> 00:09:17,505 So for example, I have this Python file which is called setup-db.py and we are trying to connect to the MindsDB. 123 00:09:17,765 --> 00:09:20,428 uh here with our Python script. 124 00:09:20,428 --> 00:09:29,187 Then we are executing SQL statement in here to create the database and the knowledge base inside mindsDB. 125 00:09:29,187 --> 00:09:33,100 We are connecting here to the Postgres database. 126 00:09:33,100 --> 00:09:38,245 We are creating the knowledge base using a model name and API key. 127 00:09:38,245 --> 00:09:40,846 You can define also the storage 128 00:09:41,428 --> 00:09:45,929 you're not defining storage, then it's going to be stored in mindsDB. 129 00:09:47,010 --> 00:09:54,034 But then if you lose the Docker image or if you do a Docker compose down, it's in the cache, it's going to disappear. 130 00:09:54,034 --> 00:09:56,195 And here, you can store it in Postgres. 131 00:09:56,195 --> 00:10:00,860 You can also here populate the Google Drive knowledge base from 132 00:10:00,860 --> 00:10:05,985 Postgres table and then we execute all this. 133 00:10:05,985 --> 00:10:08,027 So what does this do? 134 00:10:08,027 --> 00:10:13,754 as soon as I do make mindsdb and spin up my Docker image it's going to run all this. 135 00:10:13,754 --> 00:10:29,685 You can also do it here you could create everything in mindsDB in this SQL query editor here for example this is example code to connect to a Slack data source so this is to show 136 00:10:29,685 --> 00:10:30,745 you that you can 137 00:10:31,077 --> 00:10:37,391 Do it inside mindsDB using the built-in SQL editor and then creating all of this here. 138 00:10:37,391 --> 00:10:50,119 But you can also do it in a more, I would say, production-ready way, where you can have everything as code, em where you can define the API keys and have them stored in a secret 139 00:10:50,119 --> 00:10:52,080 and secure location. 140 00:10:52,080 --> 00:10:55,361 Because here you would have to write, for example, the... 141 00:10:55,381 --> 00:10:55,953 Yes. 142 00:10:55,953 --> 00:11:05,398 you paste it in here, for example, I'm creating a database for Slack and I'm gonna have to paste in my token right in there, which is not a very secure way to do it. 143 00:11:05,479 --> 00:11:07,090 But if you want to, 144 00:11:07,090 --> 00:11:14,523 take it up further for a production environment, you would have all of these secrets stored in a secure store. 145 00:11:14,523 --> 00:11:21,424 So this is to show that you can create a knowledge base, you can create an agent using the query editor of MindsDB. 146 00:11:21,424 --> 00:11:26,888 You can also, write it as code and run it when you create the image. 147 00:11:27,390 --> 00:11:30,352 I have a few questions on the three steps you showed here. 148 00:11:30,352 --> 00:11:32,854 So the first step was creating a database. 149 00:11:32,854 --> 00:11:35,956 That is then the source that you indicate, right? 150 00:11:35,956 --> 00:11:41,519 This is the source Postgres database The second step I see is creating a knowledge base. 151 00:11:41,579 --> 00:11:44,866 So you set the stage for an empty knowledge base. 152 00:11:44,866 --> 00:11:49,446 with specific columns I see and one specific content column. 153 00:11:49,446 --> 00:11:54,358 And then the third step, you do is you load the data in the knowledge base. 154 00:11:54,358 --> 00:12:03,152 I expect there embeddings will start being calculated and behind the scenes, will load everything into the knowledge base of step two. 155 00:12:03,927 --> 00:12:04,368 Okay. 156 00:12:04,368 --> 00:12:06,228 And how long does this take? 157 00:12:06,228 --> 00:12:12,461 Cause embeddings, I can imagine if it does chunking or something behind the scenes, it can take quite a while, no? 158 00:12:13,228 --> 00:12:14,748 Well, we don't have a lot of files. 159 00:12:14,748 --> 00:12:17,308 We have about 14 files or so. 160 00:12:18,028 --> 00:12:19,728 It doesn't take that long. 161 00:12:19,728 --> 00:12:23,368 We can make MindsDB here. 162 00:12:24,208 --> 00:12:36,648 And it's going to start doing a Docker Compose up and checking for the Postgres that is running and also MindsDB. 163 00:12:37,248 --> 00:12:40,408 Here you can see it's successfully connected to MindsDB. 164 00:12:40,408 --> 00:12:41,354 And then now it's... 165 00:12:41,354 --> 00:12:45,940 creating the database if it exists, of course, and voila, it's done already. 166 00:12:45,940 --> 00:12:47,442 It was quite quick. 167 00:12:47,725 --> 00:12:50,048 Okay and it reloaded these files. 168 00:12:50,048 --> 00:12:51,388 Yes. 169 00:12:52,128 --> 00:12:53,148 Yes. 170 00:12:53,168 --> 00:12:58,408 So for 14, for 15 text files, that was quite quick. 171 00:12:58,498 --> 00:12:59,218 Yeah. 172 00:12:59,167 --> 00:13:04,885 Right so now we have the knowledge base in place, it's filled up, but then we don't have the AI agent yet. 173 00:13:04,903 --> 00:13:05,686 Yeah, indeed. 174 00:13:05,686 --> 00:13:07,751 So how do you create that agent? 175 00:13:07,751 --> 00:13:11,312 You can create a new editor here. 176 00:13:11,312 --> 00:13:15,852 So what you do is that we are going to create an agent here. 177 00:13:15,852 --> 00:13:19,213 Drive, we'll name it Google Drive 2 transcript bot. 178 00:13:19,213 --> 00:13:23,153 And we can use GPT 4.1, for example. 179 00:13:24,053 --> 00:13:25,993 You define which 180 00:13:26,039 --> 00:13:33,196 data it needs to connect to, so the knowledge base, Google Drive knowledge base, and then you can give it a prompt template. 181 00:13:33,196 --> 00:13:36,780 Here I said role is what I explained before. 182 00:13:36,780 --> 00:13:46,750 You need to explain the source, summarize duplicates finish every response with certain prefix, and then concise and helpful answer. 183 00:13:46,931 --> 00:13:48,623 So you just run this. 184 00:13:48,623 --> 00:13:55,502 So you just run it, and then here you have Google Drive 2 transcript bot, and it's directly usable here. 185 00:13:55,502 --> 00:13:56,311 flexible. 186 00:13:56,311 --> 00:13:58,837 in about two minutes you build a small RAG. 187 00:13:58,983 --> 00:14:00,645 Yeah, basically. 188 00:14:00,787 --> 00:14:03,792 That's what's really nice about MindsDB. 189 00:14:09,265 --> 00:14:12,736 What we did here, the whole concept of ingesting files from a Google Drive, connecting to Postgres, this is rather a long step. 190 00:14:12,736 --> 00:14:16,601 What you could also do is here based on what is in Slack. 191 00:14:17,429 --> 00:14:21,874 So it will do some kind of some part of ingestion itself. 192 00:14:21,874 --> 00:14:24,136 And then you can store everything in MindsDB. 193 00:14:24,136 --> 00:14:26,698 You don't have to connect to Postgres database. 194 00:14:26,860 --> 00:14:27,678 So it's. 195 00:14:27,678 --> 00:14:32,698 here you have step one, create database connection that is connecting to Slack. 196 00:14:32,698 --> 00:14:35,818 this case, we consider Slack a database or a data source. 197 00:14:35,818 --> 00:14:42,078 Then step two is for yourself to just see what channels there are. 198 00:14:42,078 --> 00:14:44,779 I assume this is for development purposes. 199 00:14:44,779 --> 00:14:45,411 development. 200 00:14:45,411 --> 00:14:49,372 And here we can see, for example, 201 00:14:49,372 --> 00:14:53,192 turn Slack into a database that you can query using SQL already. 202 00:14:53,192 --> 00:14:53,773 Yeah. 203 00:14:53,773 --> 00:14:56,661 And then here we can see these are the names of... 204 00:14:56,661 --> 00:14:59,208 of the channels, yeah. 205 00:14:59,208 --> 00:15:00,157 uh 206 00:15:00,157 --> 00:15:07,558 so I want to get the messages from the learning channel here, which is this ID, so we can run it. 207 00:15:09,172 --> 00:15:12,186 And then here we have all the text from the messages here. 208 00:15:12,186 --> 00:15:17,139 We're talking about Claude. 209 00:15:17,139 --> 00:15:18,659 Yeah, indeed. 210 00:15:20,213 --> 00:15:24,108 Okay, so this gives you a way to query the data so you know what's looking right. 211 00:15:24,108 --> 00:15:29,495 And then in the next step you say, let's build a knowledge base and you provide the columns. 212 00:15:29,997 --> 00:15:31,899 Yes, you're providing the columns. 213 00:15:31,899 --> 00:15:34,341 then so here we created a knowledge base. 214 00:15:34,341 --> 00:15:36,881 It's already created, so we don't have to do it anymore. 215 00:15:36,881 --> 00:15:44,144 But then we can populate the knowledge base with the messages that we have in the channel ID that we defined as so. 216 00:15:44,764 --> 00:15:48,485 And here it's going to take a bit longer because we have... 217 00:15:48,872 --> 00:15:51,073 a lot of messages in there. 218 00:15:51,073 --> 00:15:54,478 in the back, I assume it's doing all kinds of Slack API calls. 219 00:15:54,478 --> 00:15:59,478 Yes, yes, but it's quite removed from what you're doing. 220 00:15:59,478 --> 00:16:08,598 So you can really quickly get to a point where you can ask stuff about your Slack messages and everything is really made easy. 221 00:16:08,899 --> 00:16:12,239 Yeah, it puts like a SQL interface on top of your source system. 222 00:16:12,239 --> 00:16:13,701 this is really nice. 223 00:16:14,163 --> 00:16:16,015 Okay, so the query was successful. 224 00:16:16,015 --> 00:16:19,710 So now we have a Slack knowledge base here. 225 00:16:22,423 --> 00:16:26,589 And then we can create an agent that we will call Slack Assistant. 226 00:16:26,589 --> 00:16:28,791 And we define the knowledge base here. 227 00:16:30,819 --> 00:16:33,687 So we'll call it Slack Assistant demo. 228 00:16:33,687 --> 00:16:35,354 Let's create it. 229 00:16:35,354 --> 00:16:38,921 So voila, the query was successful and... 230 00:16:38,921 --> 00:16:42,510 new agent now that is primed with this information. 231 00:16:42,575 --> 00:16:44,898 Yes, this guy, Slack Assistant demo. 232 00:16:44,898 --> 00:16:49,102 And normally now if you ask a question, it will look into the knowledge base of Slack. 233 00:16:49,102 --> 00:16:49,879 So we saw. 234 00:16:49,879 --> 00:16:52,791 ask, do you have news on Claude? 235 00:16:56,707 --> 00:16:58,068 Hopefully, yes. 236 00:16:59,089 --> 00:16:59,710 Let's see. 237 00:16:59,710 --> 00:17:04,755 So he's using the knowledge base of Slack. 238 00:17:04,755 --> 00:17:07,237 here's the latest news on cloud from the Slack conversations. 239 00:17:07,237 --> 00:17:09,219 Claude Sonnet 4.5 has been released. 240 00:17:09,219 --> 00:17:12,144 And then you have the source ID, the message and the... 241 00:17:12,144 --> 00:17:18,997 Yeah, that's probably the user identifier which you can resolve to the user so that we know who this is really nice. 242 00:17:18,997 --> 00:17:19,269 Yeah. 243 00:17:19,269 --> 00:17:23,991 what you could do, you could create an agent that has access to multiple knowledge bases. 244 00:17:23,991 --> 00:17:33,631 and there you have your problems that we had, that we discussed in the PyAirbyte video, where we saying we have a lot of different sources, and we had a lot of knowledge in 245 00:17:33,631 --> 00:17:36,331 Google Drive, in Notion, in Slack. 246 00:17:37,611 --> 00:17:40,045 And PyAirbyte was kind of a... 247 00:17:40,045 --> 00:17:44,538 parentheses to get here because we also need files from Google Drive. 248 00:17:44,618 --> 00:17:55,879 But for example, if you want to skip that whole ETL process of ingesting those files and putting them in the Postgres database where you have more control, but it takes more time 249 00:17:55,879 --> 00:18:04,636 and more management and more resources, more overhead, you can also use the MindsDB integration connector, directly connect using the API. 250 00:18:04,925 --> 00:18:08,667 creating the knowledge bases using all the tools in there. 251 00:18:08,667 --> 00:18:18,244 And really quickly, you can get to a point where you have an agent using the LLM of your choice, asking questions about multiple knowledge bases that can be all put together. 252 00:18:18,791 --> 00:18:23,696 Yeah, this is indeed quite a fast track to get to something quickly. 253 00:18:23,917 --> 00:18:26,759 And then it makes me wonder, we're running this locally. 254 00:18:27,421 --> 00:18:31,044 I would love to have this integrated into some other tool. 255 00:18:31,245 --> 00:18:40,335 So does MindsDB then also expose an API that we can use to query this agent you just created, if we don't want to use this user interface? 256 00:18:40,767 --> 00:18:44,693 because there's a top called code is that something that 257 00:18:45,734 --> 00:18:50,223 Query an agent to generate response to your questions using your connected data. 258 00:18:50,440 --> 00:18:53,823 So it does actually allow you using the SDK. 259 00:18:53,823 --> 00:19:05,014 So suppose I have my Streamlit app somewhere running and I want to have my own interface and this as a backend, not with this UI, I could actually host this somewhere and query 260 00:19:05,014 --> 00:19:06,709 the agents directly. 261 00:19:06,709 --> 00:19:08,371 Indeed, yes, you could do that. 262 00:19:08,371 --> 00:19:09,703 Okay, this is really nice. 263 00:19:09,703 --> 00:19:20,087 Of course, you're not answering the question on how you host this and how you can scale this or anything, but in principle, this is a nice way of very quickly getting from Slack, 264 00:19:20,087 --> 00:19:28,932 which is a bit locked down in the silo towards making it queryable and making this into an agent and then turning this into an API. 265 00:19:28,932 --> 00:19:31,897 I think this is a really nice value proposition. 266 00:19:31,831 --> 00:19:34,092 Maybe a quick conclusion from your side. 267 00:19:34,418 --> 00:19:36,319 it brings you to value quickly. 268 00:19:36,319 --> 00:19:41,923 with PyAirbyte for Google Drive and MindsDB for Notion and Slack, I was able to... 269 00:19:41,923 --> 00:19:47,194 query multiple sources from Dataminded in maybe one day of work or so, 270 00:19:47,194 --> 00:19:55,083 I think it's a really nice tool, MindsDB, to turn database into a knowledge base that you can query with natural language really quickly. 271 00:19:55,083 --> 00:19:56,854 It's SQL-based AI. 272 00:19:56,854 --> 00:19:58,936 It's open source, so that's really nice. 273 00:19:58,936 --> 00:20:06,342 If you have already a Postgres database where you want to connect this to, it's really quickly to get an agent to do that. 274 00:20:06,342 --> 00:20:09,497 So yeah, I think it's a really cool tool. 275 00:20:09,798 --> 00:20:10,498 Yeah. 276 00:20:10,498 --> 00:20:12,980 From my perspective, this was a really nice demo. 277 00:20:12,980 --> 00:20:15,901 I especially liked the whole Slack setup. 278 00:20:15,921 --> 00:20:20,165 And my takeaway is you can build a RAG in a few minutes. 279 00:20:20,165 --> 00:20:25,949 You get a knowledge base, you get an agent and you get an API So good for prototyping. 280 00:20:25,949 --> 00:20:26,829 Yeah. 281 00:20:27,310 --> 00:20:28,120 Right. 282 00:20:28,550 --> 00:20:31,013 Thanks a lot for the demo, Tarik. 283 00:20:31,013 --> 00:20:34,096 Looking forward to seeing more of this and to trying it out myself. 284 00:20:34,826 --> 00:20:35,853 Thank you, Jonny, for having me. 285 00:20:35,853 --> 00:20:37,511 It was nice to talk about that. 286 00:20:37,643 --> 00:20:38,654 we have some code. 287 00:20:38,654 --> 00:20:40,384 We'll put it in the comments. 288 00:20:40,584 --> 00:20:44,147 And also if you have any questions for Tarik, also put them in the comments. 289 00:20:44,147 --> 00:20:46,188 Again, thanks for watching and see you next time. 290 00:20:46,188 --> 00:20:47,188 Bye bye.