# Community Hangout, September 2024: AI Update

## Метаданные

- **Канал:** n8n
- **YouTube:** https://www.youtube.com/watch?v=334Lbexr1Js
- **Дата:** 02.10.2024
- **Длительность:** 56:24
- **Просмотры:** 977

## Описание

This month, we invited Evgeniya Sukhodolskaya from @qdrant  to present an overview of their vector store and provide a demo on setting up a RAG-based movies recommendation chatbot using Qdrant on n8n!

Max's 30-Day AI Sprint: Max has been busy this month with an incredible AI sprint! He's been working on exciting projects, and he'll be sharing his experiences, lessons learned, and the feedback he’s received. If you’ve missed it, catch up on our social media or YouTube channel.

As always, we’ll also share community and platform updates and leave plenty of time for open discussion.

00:57 Agenda
01:21 Community Updates
05:17 Product Updates
00:00 Job Updates
00:00 Movie Recommendations with Qdrant and n8n

Links:

- Sign up for future n8n Community Hangouts: https://n8n.io/community/events
- Sign up for the n8n Community Newsletter: https://n8n.io/newsletter/
- Download Jenny's Workflow: https://n8n.io/workflows/2440
- Watch Jenny's full tutorial on the creation of this workflow: https://www.youtube.com/watch?v=O5mT8M7rqQQ

## Содержание

### [0:57](https://www.youtube.com/watch?v=334Lbexr1Js&t=57s) Agenda

n so uh full program I'll start with as usual with Community updates then ol from our AI team will give you the product updates we have a couple of new jobs that we like to share with you then it's Max and then finally it's Jenny and in between each section there will be uh you'll have an opportunity to ask us questions as well after the last talk the recording will stop and we'll have an informal Community chat with everyone so let's get to it on the

### [1:21](https://www.youtube.com/watch?v=334Lbexr1Js&t=81s) Community Updates

community side um max OLG Liam and I have been in Prague last week uh for uh our first meet up there and this was really exciting uh Liam one of our ambassadors he's in the call right now he's actually from Philadelphia and he came all the way to Prague uh to join us there um so this was a very special occasion for us uh we had about 50 15 people uh show up uh it was a really great venue very nice uh nice atmosphere um we learned that about half of them were new to N1 so max stepped in and did like a three minute crash course n you please M your mic L can you mute your mic thanks and after that Liam showed an amazing project he's been building like on top of netn or inside netn I should say he's actually been changing the code to add unit testing to it um it was pretty mind-blowing also inside the team we were like very impressed with what he was doing and a full video of his presentation will be uh shared uh early next week uh and you'll be able to download uh his image as well to try it out for yourself so this is really cool stuff then Oleg had two presentations he showed uh he first showed how he connected Siri on his Mac OS desktop with a workflow so he could run workflows like have Siri explain what he saw on screen by running it through an AI agent uh and returning the information as like spoken uh content which was really cool um and then he showed A New Concept we're working on which is nodes as a tool um and he'll cover a bit of that later in his presentation today as well and Al Max gave an overview of the 30 days Sprint which was like still a bit younger at that time than today um but he has a lot of very very cool projects to share with everyone and as I said all the videos will be available next week so you can enjoy all these presentations then as well um upcoming events um our next hangout will be on October 31st and it's going to be a workflow showcase again which means you guys can submit ideas for crazy or interesting or different workflows that you've worked on and take four or five minutes to share it with the rest of the community and everyone who participates will receive one of our exclusive notebooks uh as a thank you note uh for this um you can go to n. i communityevents to see a list of these events and in the workflow showcase there's a link where you can submit your idea and then we'll get back to you about two weeks before the event or so then to other things we are going to have another Meetup in Amsterdam in early November we're still planning this but details should be available next week and uh we are also going to do our first proper hackathon in Berlin this December uh and this is still very early days but we're aiming to have like a really nice big event uh like relatively close to the office so we can get some team members and engineers and support people all join in to have a great event together um at the same time we're still um looking for help with organizing some more events and like right now we're specifically looking for London Paris and the San Francisco Bay area Paris is getting a little closer we're working with some people who are very uh very promising um London I'm still looking for someone who's passionate about NN and who has a bit of experience doing these kind of events and who enjoys like interacting with people so if that's you please head over to n. / ambassadors and uh submit an application form and I'll get back to you for a chat and see how you could help us with that it would be really great um are there already any questions here probably not right no all right um o that means you're up

### [0:00](https://www.youtube.com/watch?v=334Lbexr1Js) Movie Recommendations with Qdrant and n8n

hey everyone welcome to the September edition of the N Community Hangouts this is the ninth time we've done it this year and as we were just saying to each other in the chat these things are getting bigger and that's really exciting and uh and satisfying for us to see that so many people are interested in this um we have a bit of a special theme today we're talking about all kinds of uh AI topics um we have uh Ann engineer o talking about product updates in the AI departments we have Max our devel talking about this 30-day AI Sprints and then we have special guest eugenus Suda I hope I pronounced that right Jenny sorry about that um from quadrants giving an overview of their uh their product explaining what a factor store is uh and giving a demo of a really cool use case with n so uh full program I'll start with as usual with Community updates then ol from our AI team will give you the product updates we have a couple of new jobs that we like to share with you then it's Max and then finally it's Jenny and in between each section there will be uh you'll have an opportunity to ask us questions as well after the last talk the recording will stop and we'll have an informal Community chat with everyone so let's get to it on the community side um max OLG Liam and I have been in Prague last week uh for uh our first meet up there and this was really exciting uh Liam one of our ambassadors he's in the call right now he's actually from Philadelphia and he came all the way to Prague uh to join us there um so this was a very special occasion for us uh we had about 50 15 people uh show up uh it was a really great venue very nice uh nice atmosphere um we learned that about half of them were new to N1 so max stepped in and did like a three minute crash course n you please M your mic L can you mute your mic thanks and after that Liam showed an amazing project he's been building like on top of netn or inside netn I should say he's actually been changing the code to add unit testing to it um it was pretty mind-blowing also inside the team we were like very impressed with what he was doing and a full video of his presentation will be uh shared uh early next week uh and you'll be able to download uh his image as well to try it out for yourself so this is really cool stuff then Oleg had two presentations he showed uh he first showed how he connected Siri on his Mac OS desktop with a workflow so he could run workflows like have Siri explain what he saw on screen by running it through an AI agent uh and returning the information as like spoken uh content which was really cool um and then he showed A New Concept we're working on which is nodes as a tool um and he'll cover a bit of that later in his presentation today as well and Al Max gave an overview of the 30 days Sprint which was like still a bit younger at that time than today um but he has a lot of very very cool projects to share with everyone and as I said all the videos will be available next week so you can enjoy all these presentations then as well um upcoming events um our next hangout will be on October 31st and it's going to be a workflow showcase again which means you guys can submit ideas for crazy or interesting or different workflows that you've worked on and take four or five minutes to share it with the rest of the community and everyone who participates will receive one of our exclusive notebooks uh as a thank you note uh for this um you can go to n. i communityevents to see a list of these events and in the workflow showcase there's a link where you can submit your idea and then we'll get back to you about two weeks before the event or so then to other things we are going to have another Meetup in Amsterdam in early November we're still planning this but details should be available next week and uh we are also going to do our first proper hackathon in Berlin this December uh and this is still very early days but we're aiming to have like a really nice big event uh like relatively close to the office so we can get some team members and engineers and support people all join in to have a great event together um at the same time we're still um looking for help with organizing some more events and like right now we're specifically looking for London Paris and the San Francisco Bay area Paris is getting a little closer we're working with some people who are very uh very promising um London I'm still looking for someone who's passionate about NN and who has a bit of experience doing these kind of events and who enjoys like interacting with people so if that's you please head over to n. / ambassadors and uh submit an application form and I'll get back to you for a chat and see how you could help us with that it would be really great um are there already any questions here probably not right no all right um o that means you're up next with products updates from the AI team yeah uh hi everyone so I don't have any Co workflow to show you today uh but I as seeing as this is a Focus hangout I would like to give you a short update what's have been happening in the AI team in the Q3 and what our plans are for Q4 uh so most importantly we got two new Engineers on board JP and Eugene uh and one designer Jason uh I think Eugene and Jason are both in the chat uh in the call today so feel free to say hi uh we also at the beginning of the quarter uh we focused on simplifying uh producing some simplified use case notes as we are seeing some patterns how the basic lamp chain was us was used with the output barer but so you don't have to fiddle with setting up the output bars yourself we added these uh these three notes to basically help you achieve these use cases without that much extensive knowledge about like the L chain whatnot uh we focus also a bit on uh selfhosted added support for AMA into the tools agent and some other providers uh made it the default agent type from conversational agent uh as it should be more reliable uh we released a cell hosted AI starter kit which was a big hit uh currently it's sitting at around 2. 6 2. 7 stars at GitHub uh if you haven't seen it please go check it out it's basically a Docker compost file with all the services like AMA quadrant uh postgress and of course n already configured together and you can it both on CPU and GPU so then you can use uh you can use local AI in your na1 workflows potentially even running uh running them offline without sending any data to anywhere uh then we've improved the file upload or rather added the file upload to our chat trigger uh and merge the chat trigger with the you know this debug chat in the canvas like when you're developing a workflow so you can now upload files uh you can control which files to upload and we also handle parsing of the files to or parsing of the images to the tools agent so if you have a vision model connected to the tools agent it would automatically recognize images uh but of course you can also use it for summarization of files and so on uh and there's been many more uxix and usability improvements and buck fixes uh but our main initiatives for the quarter were simplifying the usage of the nodes as a tool uh for that we worked on a project called node as a tool which would basically allow you to connect like a subset of uh the nodes from na10 directly to the tools agent uh and being able to specify like what parameters you want for the agent to pass to the note so then you wouldn't have to fiddle with the whole uh subw workflow uh tool and it should make it much more easier to get started with these agents and to use the Integrations that uh might use uh another one another Initiative for the quarter that we progressed on was the evaluation or rather the uh execution data aggregation and this feature allows you to basically tag your past execution with the custom tags giving it thumbs up thumbs down so you can share it like between team members to like exporting them uh and in the future it would also be available for uh evaluation and that's bringing me to our plan for Q4 you go to the next slide art uh so we're going to continue with these uh main uh initiatives releasing the note as a tool and also working on the evaluation uh we want you to allow to be able to create like a test workflows and to be able to specify criteria and to see like how your workflow tests are passing failing um so you can do that you'll be able to do that from na1 we're also starting to seeing some common rack use cases and want to simplify that as well I'm not going to go into too much details about this one yet as still the specs are pending but it's definitely you can expect something in this quarter uh also simplify tools usage with the AI agents with the both the nodes as a tool uh but also we want to improve how the support flow tool Works uh giving you a bit more visibility over uh the kind of parameters that your agent passes and make it more reliable and finally we want to also make some more chat uh improvements allow you to render more complicated chat elements and be able to control flow the flow there uh last Point here is that we are actively looking for people to talk to regarding this these AI features so we can learn how you are using them maybe what's missing in them what you would like to see uh or things that you're struggling with uh so if you would be interested to having a short user interview uh please feel free to reach out on this uh link I'm also going to post them in the chat in a bit and yeah we'll get in touch um and that's it for me thank you very much very nice thanks o um and yeah I think the link is coming up um you kind of predicted there would not be many questions and I was keeping an eye on the chat and you were right there was nothing specific about this questions about the AI starter kit um but they've been answered in chat already so that means we can continue There is one question but I answer I can fully answer it's how might someone get the basically the selfhosted AI starter if they've already installed today you know and it and separately well depends how you install it if you're running it via Docker then you can just copy some of these uh some of the directives from the docker compos file in that starter kit uh make sure you connect the network or specifi the network directive to make that both all the containers are in the same network and you should be able to v PR for back up yourb first yeah that's a good point backing up is always a good idea yeah was that all Max are there any other questions that's how I saw those an answer all right okay well guys if we missed something let us know and we'll get back to it later right um thanks Al great update exciting stuff coming up in Q4 then very quick update on the jobs front um as always we have a list of jobs but our hiring team asked me to put a spotlight on these two roles we're looking for a senior product manager on the integration site so that means you're going to be working with the nodes team helping make sure we have the right focus and we understand the needs of our users and the uh goto Market team is looking for a senior data analyst to help us crunch the numbers and see the impact of the work that we're doing um we have more jobs so have a look on N. C careers if you want to know more there's also a lot of information there about our company culture and how we work together how we work remote uh throughout Europe Etc very good stuff um and with that uh it is time for you Max um what do you want to share your screen um I can speak for a second then I'll share my screen in a moment okay cool sharing um hey everyone so good to see you all um so some of you that are sort of deep in the NN Community might know what I'm about to talk about um but I've been doing a 30-day AI Sprint for the last 25 days now it's day 25 and I'll just give a little context as to you know why am I doing this so I transitioned into devell at NN from design um about 68 weeks ago it's been sort of a transition period um and one of the things I realized that we need to do more of is content that shows people how to use NN but also inspires I think uh in the content I've been doing today there was plenty of power users that didn't even know about the form trigger for example you know and at team at end that's table Stakes we all sort of know about the form trigger and that's just one example so it's about getting the word out and I think showing people what's feasible today with AI and then what also just the learning from doing this stuff you know in production and building it so that's what I've been doing for the last 25 days it's been a whirlwind Journey uh not so much sleep honestly it's all self-imposed I just can't stop building it's very difficult to and we've launched a ton of projects um and I'll share my screen and go over a few of those projects but everything I've built it's all freely available you clone duplicate change um you know I've gotten a few people that have critiqued some of the flows that I built and rightfully so you know some of these were sort of knocked out and put up and I think the thing about building imperfect um with n ATN shows firstly that you can ship something imperfect that creates value um as an MVP as a minif fiable product and then that it's possible to improve it and make it better for some use cases that's going to be before you launch the production because it's very important and for other use cases like an internal op something um you might launch that sooner but um being able to show people what that looks like when you're actually building today has been I think really rewarding a lot of people have been really appreciating it so let me I just send a request yeah there you go yeah Max you've also been working with a couple of community members on these projects right that's right so that's what I can go over in a second when I um I show off all the things that we've been doing here I do have about one million tab so you just give me one second here I'll find that and I believe you're about to see corck tab yeah so project dour this is a project we actually launched today I built this along with Oscar from the from our community runs an amazing Channel where he makes great YouTube videos um about NN and he's also professional automator so for me it was really fun working with him because I think it's our community members that often know a lot of the you know serious NN uh production you know skills more than some of us so we launched this today it's on product hunt um and what Email Spy does if I visit it scrapes various different websites uh for a given search domain and outputs and emails if I search for NN here this should work in about a couple seconds CU I've already ran this query before and we're caching this um but this entire experience that you're seeing here you see this there's three emails we found and we've seen the different websites we found these on so here we're finding out on the NN website on crunch base um and how this workflow works for example its backend is all powered by this workflow here now I'm not going to go into the details of each workflow um but I do have a vlog for the AI Sprint where I do go into those details and in the next one I recorded a session with Oscar who went over a lot of the details of this Oscar did a lot of the heavy moving parts of this workflow but in a nutshell what we're doing is we take in the uh that domain name uh we performed some searches using Brave search and then we have a few different methods where we're analyzing that so one method is using Puppeteer this is a home roll solution Oscar built and another meth Max sorry for interrupting could you zoom in a little bit to make the readability better sure thing and um there's two other paths that basically use different AI methods in combination of traditional scraping to Output um whatever emails we can find for that given domain and then we merge that all together and you duplicate that sort of more standard ETL and then send it to the front end now again if you like the details if of this or if you want to rebuild this um project it's all available in the 30-day AI Sprint um notion homepage um so in here there's a little details on the project the various projects I've launched um links to the Vlog and then also if we scroll down here this is the single source of Truth for all the projects I'm building so there's email spite that's the one I just showed you today sty ens is GitHub profile analyzer very cool stuff we've got to research paper summary assistant um and a bunch of other stuff as well um but for email py that's the project that we're launching today and I think it's the most complex one we've done so far and it really shows some moving Parts it shows um using AI agents in what I would say is a production use case you know I think a lot of folks if they added a few more bills and whistles to this could be a paid SAS for certain roles we're giving it away totally free because this is just an inductive tool for you guys to understand how we built that um some of the cool details about a project like this is that you might not see in like introductory content is you know this workflow could fail at any uh one of these steps right it's possible that happens in production uh the way Oscar set it up is the error workflow receives the call back from this workflow and so it will send it to the front end so if you're using it today you know it may fail a few percent of the time because again we're doing web scraping of dozens of sites and you know we put this together in a few days but the nice thing is there's a flexibility in idend to for example serve the user uh an error message now I was trying to conjure that error message to show you that it's a little too robust we've been improving it throughout the day so it's kind of hard to do um but yeah go check this out and um tell me what could be better about it and I challenge you to clone it and duplicate it and use it yourself uh a lot of the pattern in these things even if you're not doing that specific use case you could apply it to your more serious business use cases um so yeah I've got four more days with the AI Sprint um with Marcel one of our community members on the weekend we're doing a hackathon and we're going to be building a data assistant that knows statistics and these sorts of things and can crunch data analytics fingers crossed that's definitely going to be the most complicated thing we've built um but before I hand off back to bot I just want to say I'd love if you guys follow along and get inspired to build if you do build something send it to me and since we did launch on product hunt today now I'm not allowed to ask you to upload it but I'm going to share this link here what would be great is if you could check it out and then show your support if you like it thanks Bob brilliant Max thank you um let me just quickly check if there's any questions could you share the link to your email spy here Max people are asking for it oh I most certainly can and Max is our new Dev here franisco was asking that yes can you say a bit more about that role Max Maybe find interesting yeah so I'm I guess my title is senior developer Advocate but we can kind of throw that away because I think what my role really is at NN is a bit of a resident flow gramar um you know we have bot focusing on community um and that is a lot of different things right events and everything so I get to focus a bit more on the subset of the community basically help evangelize um them building unblocking them building a and creating educational content I think a lot of us are in dep I Learners I think for a tool like NM it's a lot better to see something that's working that's creating value than you know read a novel so that's the kind of stuff that I want to create and also I think a lot of the stuff is already happening in the community so be a catalyst for that tell the community stories every Meetup I go to there's so many fascinating things all of you guys are working on I also want to as we ramp up help tell that story so focus on the Sprint right now we're going to wrap that up and basically do a retro and figure out how we do this in a more sustainable way CU because I would probably die if we continued in the Sprint format but yeah that's what my role is and exactly what it's going to look like might change over the next months but it's basically using NN showing people how fantastic it is and showing I think specifically how it works in prod and the learnings from that CU I think it's our community that actually knows the best on those things super cool thanks Max and so yeah from one developer aspect to the next our next speaker is Jenny suaya from quadrants uh she's developer Advocate there as well and Jenny will briefly remind us what semantic search is and show us what a vector search data bear database like quadrant can provide Beyond semantic search she will demonstrate how to build an build in n8n a rag recommendation chat bot and why it is better than asking GPT to recommend something for you so you ready to be amazed Jenny yeah uh firstly Bart Kudos on the second time your surname pronunciation of mine was Chef kiss but so I would consider it the best start and uh actually working back to Max uh I had a couple of times uh also explaining to people what is developer Advocate and I think I developed my own formula I call uh developer Advocate a standup comedian in it so like when you take the both and you just combine it in one so I'm here to entertain and uh I'm going to do exactly what BART read and I hope after all this cool things with the ice print I won't make you disappointed because honestly I am looking at Max's job with a little bit of fear because he setting the you know the plank for others here and I want my slep so guys you don't do that to your colleagues in the field um so I'm going to firstly give throw in a very little slide explaining what is Vector databases I'm pretty sure you all guys already know what is that but it's just a general reminder um if you don't know it's going to be clearer and then I'm going to show a little demo and I mean let's see how it's going to go but I hope it's going to work out and you will like what we built yesterday with ALG and also I want to separately say kudos to Al he wanted to call me yesterday for 30 minutes to go over the case we spent in the end one hour 30 minutes talking about bugs in the both sides and I loved it honestly so much uh I think we did something very fun so uh about Vector databases uh so what are we talking here about I guess just defaults a good okay so the first thing that you have to start with explaining what is Vector databas is about embeddings and embeddings is simply vector representation of all data that surrounds you it can be textual it can be image it can be audio it can be video it can be anything it's just one of the ways to present your data in machine readable format that you can basically perform some mathematical operations on it for example the operation of similarity so in this space of representations some things which example like text would be similar say cheese and GAA it's semantically similar thing so when present that has neural embeddings they're going to be pretty closed and this is exactly where we get this representations from we get them from naral networks which are like now super big thing I think basically you can't find a person who doesn't know about them and uh this is happening like this representations are possible because they see for example if we're talking about text they see so much text that at some point they develop this pattern under understanding and see that some words are belonging together and they're more similar and some words are like more far apart and in this numerical representation Vector representation we can then perform different operations and uh I mean here is an example like you know no words overlap no whatsoever but it's very very simple sentences to us and why what in this picture doing Vector databases where they are about searching this information this encoded mathematically information in a very big space of this representations because if you can imagine if we are working with data at some point we have a grandio like um the amount of points is humongous the amount of data that we want to see to compare its similarity find out what is similar find out like what is there and for that we need a specific instrument so it was the same for relational databases which stored the data in a specific format but with this embeddings which are coming from uh neural network work we need a specific tool and this are vector databases because they allow to store this Vector representations of documents that you are searching of sentences of text of images of videos at a big big scale which happens usually when you load the data you encod the data you store it and then at the inference moment with Vector databases you just ask some question it's also converted to this mathematical Vector representation and then in this huge space that we have behind stored in the database huge space of points and representations we find a similar candidate I am going to actually show you when I'm going to be demoing a representation of films which are going to be in called this mathematical vectors and shown as a points uh how like our database um combines them together uh behind like under the hood so H why they're so big right now this Vector data basis and I think now I'm going to say the um the word which now people either hate or love it's literally rug it's like retrieval augmented generation and I think that was the Catalyst of the popularity of the Spector databases because they are used under the hood with these models because when these models arised like two three years ago we saw their potential we were like wow finally we can throw away Google or whatever is that and ask any question in a simple form to this model and it's going to provide me with an answer and I'm going to live in a perfect word where I will know everything but then you start noticing that when you provided questions it start sometimes generating some very strange stuff which is not true or which is absolutely like not related to reality it never happened and you think like what is that and that is hallucinations because even the biggest model that was trained on the biggest amount of data from internet is limited to this training data and we can't trust it with answering our question correctly without providing to it some context some external context and the perfect thing for storing that context is actually a vector database which allows under the hood to perform the fast search for something that we want to find out for example be it a documentation we ask some question it's not keys I am going to ask like how to build um I don't know how to build a search in quadrant how to install an A10 AI hosted kit or whatever if I'm going to just search it in Google with keywords it's not going to work perfectly because who knows if in documentation there are like the same words that you are formulating the question but uh if you are going to convert it to mathematical representation and use it in the vector database it's going to find you a similar answer similar chunk of the documentation it's going to give it to our model and it's going to generate a beautiful answer related to documentation for example and then we're going to see basically this powerful instrument which allows us to have a search engine to which we don't have to specifically formulate some questions like thinking about the keyword or thinking about how to ask something and uh not getting any answer but we can just talk with it like with a chatbot and also it's not limited only to text it also can work with images with videos with audios because everything is basically can be represented in this Vector form so I from Quadrant and quadrant is one of the vector databases which is using one of the latest algorithms for that written in Rust and obviously is the best because why I'm here if not to advertise it but the thing is that beyond the classical search API which some people think about semantic search and uh they say okay you can do you can compare the vectors wow is there anything else I'm used to use my relational databases with this equal this filtering an order by we also have that we can provide nested filters and different types of aggregating information not only semantically but exactly and all the types of combinations of that which like you know when you can take the both like the best of both cases it's always the perfect solution but I guess one of the interesting features that we have and that what I'm going to demonstrate today is recommendation app and it works on the pain that some people when they search for something they not only want something specific they also don't want something specific for example like as I'm going to present today I'm going to do a super small short movie recommander system sometimes I want to watch some film which is about friendship but I don't want to see anything about war or pain or blood because you know what I'm not in the mood 2024 is already a year you know so that's what can be done behind in our knowledge based in quadrant and today is exactly what I want to demo you I want to show you on quadrant on a free cluster that we have so if you want to replicate it and do it with your own data also try it that would be nice I want to show you a demo of how to create a simple movie recommendation agent where you just write in the chat hey I'm sitting home I want to watch something today about love but without vampires because after Twilight I'm traumatized for already 10 years and it's going to solve your longing and not only it but you also can do it just in na10 and without coding which honestly I loved a lot because I had a lot of fun yesterday trying to build that around and that's what's gonna be about I don't want to do this demo also from scratch of course this workflow is not as big as Max showed for his AI Sprint but still like I mean look at this sausage I think I'm going to do a detailed video I will use the experience of Max and I'm going to do detailed video I say tomorrow doing it from scratch but now let's just look at the components and let's just see what it can do and in general like understand what is happening so before I am going to show that it's actually working as it intended to work so as a movie recommendation chat rug system uh I want to show the first part which is uploading the data to quadrant blog quadrant Vector store because if we want to create a movie recommendation system we want in our database to store the movie these descriptions so they're going to be retrieved and then recommended to the user and uh of course it doesn't have to be movies but it don't in our particular case I took them and uh based on that our model our agent is going to do all the recommendations without ability to invent some movies which don't exist for example so um to upload data to quadrant uh we usually need like our cluster I have cluster it's actually free tier because I don't know it's enough and I already have created a small uh so I'm going to be using this data set I took it from kaggle and uh honestly it's a good choice if you want to showcase something it's top thousand IMDb movies with their short description so you can see what is there and um to save your time I'm not going to upload all 1,000 points now because as you understand they have to be converted to mathematical representation so two embeddings in our case with open AI model but you can take any open source model because guys have here also like hugging face and they also have open Llama if I am not mistaken yeah oh llama too um so to make it more fast I just want to in general firstly show you how to do it for smaller amount of points so what is happening here we're downloading my data set from GitHub we're extracting the file which is pretty similar and then let me limit um the amount of points amount of vectors uploaded to the database to 50 for example should be enough for the demonstration and Y and here we're going to since this is the one of the containers that we are already saving for a bigger demonstration let me create a smaller container here for 50 vectors and see how it's going to go and I am going to run this so this is going to upload our representations made by open AI uh embeddings to our database and it's going to uh take from the data from the so we can see it from here we get the file in the file we have all of this movies The Classical form and so here it's going to take all of this information about movies and it's going to encode in mathematical like representation only descriptions because user usually asking for some movie recommendation is going to ask something around description he's going to say okay give me something about like love but not much about like sweetness School romance and that's what we want to basically search around and let's test this stab fastly and as you can see they're uploading one by one to our database and let me show how does it look here so for each cluster we have a dashboard and this dashboard shows already points uploaded and I can show for our big collection so this smaller collection is created now you can see that now it has 20 vectors it's going to be more when it's going to finish uploading but let's look at already done uploaded big collection of 1,000 movies so you can see how they are stored we have all the meta data uh about them we have a Content uh and we have a vector representation which is created through the open AI embeddings which is 1,536 points and we can see them on the graph how they are looking and are they close or not so we have all the different fun to play around and for example let me see in the visual visualize all the points that we have in our films and for example we see something very close on the graph and it is here Harry Potter and prisoner of acaban and the point next to it is Harry Potter and the Sorcerer Stone so they are grouped by description and you can see that the points similar to each other are about the similar films for example here is Star Wars and Star Wars so the same way when we are going to ask for the recommendation we our database is going to find a point similar to some of the films that are already existing and point as further as possible from the negative recommendation of the user so we can see all the 50 items uploaded and we can already check it in our collection so that's how it's done in general and now I would love to demonstrate you what we did yesterday with the i engine because I think it's like super cool firstly I want to literally show how does it work and then briefly go over it so there is a chart and uh this is a movie recommender agent and I want some to watch some film which is about sport but not about for example and then we will see the recommendation being provided to us which is going in the meanwhile to call the tool and it's call in the oh I love this demo effect it's when you're doing something for 1 million times and then it works perfectly and then the demo comes and then nothing comes let me try again so I want to film about sport but please nothing about be uhhuh and we got our answer and our answer is sorry it was supposed to be in the chat yes so the wrestler The Story of an Indian athlete and champion and the story of a dog which is kind of Sportsman but not much we can see that it called the tool which we built here this tool is basically our workflow and this tool is called in our database and our recommendation I which returns top three films from our database with the score of similarity as you can see it's not that high honestly speaking I guess because out of 1,000 films we don't have so many films close about like Sportsmen but not baseballers but actually two top two answers are pretty good so it's about the professional wrestler and it's about the story of an Indian athlete and nothing about baseball H so here our agent is just talking to uh open AI model so it's like GPT just is used to call the tool and it used to basically beautify the answer of the tool the retriever data from the database and what is happening here so here we're firstly taking our positive and negative example because our um our model is smart enough to provide us based on our request a positive and a negative example so the positive EX example the film about sports and the negative example is nothing related to baseball and then uh we embed these examples in the same Vector representation as we did here because we need to compare same vectors it makes sense you compare something similar to something similar then is basic mapping of the points where taking our embeddings our Vector representations and there we are sending uh them to quadrant recommendation API which basically looks like this you provide a positive example something that should be close and a negative far and some strategy of aggregation after that quadrant returns as points uh so what is recommended you can see here I limited it to three points so it recommended top three points which are the closest we fetch them from database because we want to know what they are about and you can see that it's fetching all the descriptions uh like about the wrestler the name the film and yada and then we just beautifully combine them and then we send them to agent and that's how our agent can basically build their answer on something which is like truly our document in the database and I guess the cool thing is that you can do it not with just the movie data set which is pretty easy but you can do it with your own documentation for example or with your own product or with Amazon or with house searching and then you are just going to say okay uh could you please recommend me something say I'm searching for a house for renting a house like house with through rooms but please without like red wall colors because they make me psyched or something and it's going to be done with this workflow without coding in one click and honestly that's sick and I love it so I am super grateful for the ability not only to present but literally to just try the tool because that was so much fun combining this simple solution brilliant Jenny and I know how hard it used to be to build recommendation engines and to see you do it in such a short time is it's amazing hly it's like kudos to you guys I mean also called as quadrant what can I say exactly it's a it's beautiful combination and I'm reading the chat and people are super impressed with your tools and the visualization that the demonstration of how the movies were close together in the vector space even though you're like flattening what is it 12 dimensional space to two I imagine yeah it's it's simplification of course it's like uh if I would be mathematical it's like principal component analysis but yeah it's like basically flatting into one dimension but it's still as you see it grouped Harry Potters together so it makes sense yeah absolutely that's really yeah and I think uh yeah people are saying we didn't see that in other products yet so that is really cool um yeah I'm reading a lot of Praise I don't think we have a lot of questions unless someone has one right now uh who wants to share one oh I noticed my uh CEO in the comments so all the questions can go directly great oh hang on now there are some questions here um so Marcus asks how would I keep the embedded documentation up to date in a vector de database I would want to have help articles updated and added and need to embed them do I once a day or could I make sure the docs are always in the vector space so basically do you need to like re-encode everything all the time or can you do you have a smarter strategy for that no I mean um it depends on the how often your data is changing because like the new data arrives you just encod a new part of it and add it like you would upload any data to database right if you want to store it there you just upload something new but you don't have to rein code the old ones because they are already in the right representation there if something changes you will have to re-upload it just because it makes sense like for example I was storing some say we had like a guideline we had our documentation then it completely changed uh you of course want to reupload it because the old data is saved but it's just a common sense if your data is changing a lot I would definitely do a day Shadow three applo it's fast because uh I mean quadrant is built on Rust all the uh encoding is done like we have Vector stores which is up to 520 million points and it's working pretty fast because we have horizontal and vertical scaling so it should be like a piece of cake but it's good to have a strategy in place to minimize the encoding costs right yeah of course so that's okay that makes sense um another question is from oh hang on I just scroll out of view uh Colleen she said this was great and inspiring I started looking at quadrant the other day and wasn't sure where to start will this be an available template that we can dive in and look at so if you could share this with us I can make sure it get published as well yeah I would be happy to I also was planning to do tomorrow the whole video of setting this app from scratch but uh you guys need to teach me how to share na10 workflows because I think I am kindly using kindly provided Ole account so he has all of my data we have to workflow Library where we can publish these yes yeah but you know what it's a very interesting uh thought in a comment to us that the person didn't know where to start from and we are going to like I would be happy if somebody reaches out to me and says what is the problem with starting because that's exactly the number one thing that you would like for your product that the people feel like they want to start I think recently we have doing a big job of moving very easy quick to start tutorials to our Cloud so maybe it's the nicest way would be to start like sign up in the cloud and try to just click around without even reading the documentation because Co reads documentation you just I was with n to end I was like I'm going to figure it out but you had an O right that helps yeah mean o was like documentation yeah this is really cool let's make sure this get published so when we share these videos there will be links below them to uh to Jenny's full video as well as to the template so you can download it and get started right away and let's make sure to include your your voucher code as well for that free one gigabyte clust oh yeah it's it's just for everybody everybody you can sign up um let me see another question is from Carter Harrison and he asks what makes quadrant different from other Vector dat databases like pine cone oh it's a oftenly asked question I would say my personal opinion the thing is like I've heard a lot of different uh answers based on like who is we have four de RS we have four different answers and everything I would say for me personally except like the scalability in the thing would be firstly the uh the reaction on how fast we develop the features reacting to the community if you look at the quadrant Community it's very responsive very open sourced and very managed so our CTO is sitting there daily answering to all the questions because he just loves his job honestly speaking so for example this visualization was a thing that everybody wanted and we pretty fastly did it and so except from just being a good decent product for Vector search which is already around for four years so it was before the old hype and it wasn't done on the knee like you know behind the scenes behind the llms thinking faster we want to earn our money so it's not only like a stable product it's a product which adapts to what their Community needs very fast in a true open source spirit this company because I liked how the community works thank you yeah um let's see I have a couple more we only have a few minutes left so I'm going to pick a few um so that Mark crio ask is there any way of identifying datas of the data Deltas of the data so that only new is added I think that's kind of answered already uh Mark that that's on the NN side you need to like figure out which data is new right yeah so we don't have like a standard process for that um by the way just a quick tip here in we have a compareed data set node which would make this easier because you just connect two inputs basically and it would give you like a difference between data based on some criteria so that might be a case cool thank you um Jim Lee asks when will we get an official quadrant n8n node still too many steps need to manually generate the embeddings for the recommended API would love to only just have to supply the text instead well that's not to me I think is it on a road map I can't promise I framed you leave with it no h and I actually when you reiterate with the question about checking if the data is same or not I realized what you were asking about uh because I interpreted it wrongly and of course in quadrant you can check if there is already some data because we have payLo which is the fields which you can filter with so when you insert you can just addition basically the filter how to say condition thank you I'm sorry English is not my mother Tong so you can add a filter condition and uh that will help you not to insert the same data because it's going to be already detected that it's in database yeah for some reason I was trying to understand like if the data is being already saved or if it's F I don't know what how I understand the question first time was too excited after the demo and also I saw somebody asking the question about PDF documents and if quadrant can work with them every data uh you just need to find a way to convert it to Vector representation and with na1 for example you can easily do that we did that with uh o yesterday you just upload a PDF document you split it in chunks and you upload it to quadrant and it takes like what three blocks of an ATM four okay maybe four I will give you that um last question because we're running out of time here um Alex asked should we ask a quadrant note with insert documents for the first pass as shown as the demo and a separate workflow with a call to quadrant points API for aborts to account for new source document versions not sure I like that's more an n8n question Alexia that like your initial upload versus processing the Deltas right yes it was just a question if we had a PDF that went through and then that was version someone made some updates could I overwrite the points in my collection for the old document so when I'm chatting with it I'm not referencing the old dock I've managed to go in and remove those points so my Vector store is always accurate firstly I'm trying to figure out is it questions still to me or to BR I guess that's question to us uh so we currently don't support the upsert in uh operation for quadrants we support for some other Vector stores but we didn't add it for quadron yet so you would have to check it yourself based on this metadata see if this already there andul yeah I guess it's using API would be now the same with recommendation API I mean guys faster so many requests for Quadrant groups I am um okay so there there's actually more questions here but we're running out of time so I'm going to wrap up and um let me just quickly go back to my slides hang on um if you have other questions guys please head over to the Forum on community. nn. and open a topic there and we can continue the conversation there um so for now uh all I want to say is thanks for joining us here thanks Jenny for your fantastic presentation I really enjoy the example and the ease of achieving such an impressive result um we hope to see you again at some point in the future this was really nice our next hangout is going to be on October 31st it's going to be a workflow showcase as I mentioned earlier if you want to participate and win uh an awesome NN notebook head over to nn. communityevents and you can submit your idea there and for now that's it if there's anything else we can do for you please email community at nend. io that will go straight to me and we will get back in touch with you as soon as we can

---
*Источник: https://ekstraktznaniy.ru/video/15587*