Let’s Build a Movie Recommendations Next.js App with Vector Embeddings (Generative AI for beginners)

[@SonnySangha]

UPDATE ON IMPORT BUG: There is a known bug when importing, it currently works for 50% of users - Datastax have been informed and are working on a fix

[@LongBoy.0]

I wonder if it has something to do with Rate limiting coming from OpenAI? at least for me, that sounds like it was it. It looks like being in Usage Tier 1 is possibly a reason why some are getting rate limited. Just guessing. EDIT: oh weird. I was getting the 429 error after deleting the collection and retrying, over and over, then eventually I gave up and came back 10 mins later and the database went from 0 records with the error banner to 250 records and no errors. 🤷🏻‍♂

[@Selfnrj]

Getting the 429 error too. Hope for a fix soon!

[@ihsaneakhmoul]

​@LongBoy.0 hi did it work yet ? It did not for me

[@LongBoy.0]

@@ihsaneakhmoul yes, it has been working for me and I finished the project. With Astra DB, I'm not sure why it suddenly started working. I got an error code that was for "rate limiting" so I went to my openAI account and desposited $10. Shortly thereafter, I recreated the DB and it worked. I'm not sure the money had anyhting to do with it though, since my balance still shows as $10 after running the project DB for a couple of days.

[@bineetjha2268]

@@LongBoy.0 is their any alt option to resolve that bug 'error'

[@javascriptexpanse]

Known this channel from 48k subs and i have to say this is the only video that got my attention - liked, commented and shared. more of this sangha

[@albertfranklin2317]

You are an amazing teacher my friend. Keep up the good work.

[@SonnySangha]

Appreciate you!

[@zahrahusain9468]

Dude has outdone himself yet again Explaination Very clear and on point Sonny this was incredibly helpful....... Keep 'em coming..... Please... Please....... Please

[@SonnySangha]

You’re welcome!! I’m glad you liked the explanation! Will keep em coming🔥

[@hiteshnalamwar2722]

Great Info we can learn. Love to see videos on AI agent :)

[@developmentwithtariq]

Thank you for your great contribution in my life as a developer

[@100AFRO]

yessir!! thanks papaReact. let's dive into the ai era..

[@OsamaZaidiah]

Thank you so much Sonny

[@Diegoalllll]

9:17 we want it

[@thatgitahi]

Of course I'd love t see an AI agent video on the channel. I think it would really help me on a project Im working on

[@johny962]

Thank you very much for the great content! It's very useful. Keep it up!

[@tedreams]

Good content! Thank you for getting straight to the point.

[@SonnySangha]

I’m glad you liked it!!

[@RashiAgrawal-r1t]

Hey @sonny sanga any update on dataset loading please or you can provide alternate dataset?

[@jonosuav]

Awesome content! Would be interested on seeing how to further extend it based upon user likes

[@mileslegend]

Was missing you bro

[@ad5555]

Thanks for great video! I also get problem with uploading the sample data and loading the dataset. Have anyone found a solution?

[@SonnySangha]

Working on a fix with their team you are correct their is a bug on their side!

[@devotee6586]

bro, I am facing an issue with my data not loading It gives me an error:- failed to load

[@Faizanakyt]

Me too :(

[@devotee6586]

@@Faizanakyt still facing same problem?

[@Faizanakyt]

@@devotee6586 Yup

[@Faizanakyt]

@@devotee6586 Does it work for you?

[@LongBoy.0]

I just got this issue, I think it's because the 512 vector is too pricy for the free tier. I think you need to lower to 256 or just pay a little $$ to make the project work.

[@shimronshitrit5187]

Hey Sonny, I watched your video on building a movie recommendation app with vector embeddings, and I found it really interesting! I'm currently working on a project where I need to match users based on their job titles or industry types. Do you think the approach with vector embeddings that you showed could be adapted for this kind of user matching? Any advice would be greatly appreciated!

[@larsh5853]

Very cool. Question: Lets say I have a database with movies and categories on which each movie is rated. Example categories could be What would be Plot Originality, Character Development, Acting Quality, Tone and Atmosphere, Music and Sound, etc. And each category is rated on a scale (such as 1 to 10) or given a qualitative assessment (such as "Excellent," "Good," "Fair," "Poor"). These scores combined provide an overall rating for each movie. How would you apply the principles in your video to develop a Movie Recommendation app where users start by filling out a brief profile where they input their preferences across the different categories. For example, they might be asked, "How important is acting quality to you?" on a scale from 1 to 10. So, I guess the recommendation algorithm should focus more on matching user-defined criteria with the multi-dimensional ratings of movies? Is that something DataStax & Astra DB can support?

[@LongBoy.0]

Love this content.

[@mileslegend]

And your content 🥳🏅

[@SonnySangha]

I’m back!

[@mileslegend]

@@SonnySangha could we build something to automate tasks...

[@SonnySangha]

I like the idea of

[@SanZLan]

Great video. Thanks. I would like to hear more about how the personalized recommendations could work. Would this work by creating a single vector embedding from the set of contents that the user likes and searching by that "grouped" vector embedding? Is the functionality for that also included in Astra DB and/or through the Open AI API? Or were you referring to a search where you show some recommendations for each one of the liked contents? If these questions are too much for a quick reply, this would surely make for an excellent follow-up video. Anyway, thanks for the information. Keep up with the good work.

[@SonnySangha]

Love this was thinking the same thing - will get onto this soon

[@jamestuttle]

+1 on you building an AI Agent :)

[@TheGarageboyz]

Great video! Thanks! Netflix recommendations (in my case) are too narrow imo. Question. Is Next JS better than React?

[@SunnyAnsari]

Next JS is a framework for React, that supports the server side React component.

[@SonnySangha]

@@SunnyAnsarigreat reply! Yes I recommend using Next.js with React always!

[@apokalupsis116]

Do video on how you always have the best lighting in your videos

[@SonnySangha]

Hahaha gotcha!

[@AloBara-b1b]

yes please build AI agent video, it be interesting.

[@hashblast3r]

Any good sites to get free csv datasets? Having trouble finding some. Keep up the great work buddy. This was a fun one

[@bineetjha2268]

kaggle

[@hungxnguyn]

15:15 dataset isn't able to load to completion, even after multiple attempts, and refreshing the page. Anyone find a working solution?

[@MessiLeo2312]

same problem dataset not load

[@pushkarshukla5340]

did you find a solution to it

[@huzaifakiani9059]

@@pushkarshukla5340 you have to remove the "/" from the columns header and replace it with "_". For e.x replace "Ratings/0/Source" with "Ratings_0_Source"

[@callmekelvin]

Do we need a paid version of Openai, because I am not even able to create a collection, just says Your collection is ready to use but it's not showing in the collections section 😕😞

[@SonnySangha]

You should get abit of free trial credit when you sign up for openAI for the first time

[@herbertk9266]

Thank you 😊

[@pumpkin2228]

AI Agent as well 👑👑

[@SonnySangha]

Awesome - will look into it

[@mileslegend]

hey bro, it seems my data set is not loading.... i dont know why its behaving like that, unfortunately i cant follow along i dont know the cause of it,

[@LongBoy.0]

It's because the free tier isn't enough to get this project to load. Try 256 or just pay a little $$ to Astra to make it go :)

[@ramprasadchauhan7]

Nice explanation, please make project with trained ai chatbot model using langchain js, . Thank you

[@mileslegend]

lets see the AI agent being built

[@divyv20]

Hey Sonny , very good video . I can do better editing in your videos which can help you to get more engagement in your videos . Pls lmk what do you think ?

[@Tarun8336]

Awesome

[@DSoundTrackSaga]

🤘🤘🤘

[@pumpkin2228]

Please make using React as well ❤❤

[@SonnySangha]

You mean React without Next? Because Next is a React framework! Not sure if that was a request or confusion

[@pumpkin2228]

​@@SonnySangha yes without Next js please

[@SonnySangha]

But why…

[@caiohportella]

My dataset never finished loading for some reason...

[@shadeis9276]

Same, the dataset doesn't finish loading and I'm stuck with 0 records

[@caiohportella]

Literally the same, bud!

[@shadeis9276]

@@caiohportella I think the problem is with the OpenAI key. Sonny probably uses a higher tier key.

[@caiohportella]

​@@shadeis9276 hmm makes sense, do you know if adding a credit card to OpenAI will charge lightly per use? Is it worth it for portfolio means?

[@shadeis9276]

@@caiohportella I have no clue sorry, this is like my second time trying to use OpenAI keys. Though on the DataStax integrations section, there are a bunch of other embedding providers than OpenAI.

[@homemmacacoverdadeiro]

My man.

[@usmanmarkaz]

bro please one full stack SaaS app with React native and AI please

[@SonnySangha]

I hear you

[@usmanmarkaz]

@@SonnySangha thanks thanks a lot

[@LongBoy.0]

Please build an AI agent for us, that would be sweet

[@SonnySangha]

I got you!

[@LongBoy.0]

@@SonnySangha Yo quick question - having a bug. From the below code in the movie[id] route, the movie object that gets returned in line 25 does not have a property called "$vector" how do we find similar movies? const similarMovies = (await movies .find( {}, { vector: movie.$vector, //

[@ananthegde2107]

Hey can you make a video on MERN stack projects like Stock trading application or Uber clone application

[@blacksun6761]

Indians what is you problem with mern?!

[@ananthegde2107]

@@blacksun6761 can you be specific

[@SonnySangha]

I have to agree, all MERN is is express & mongoDB with what we do - I have taught this loads!!! You also don’t need express when handling Next.js route handlers, stop following the trend of saying MERN stack and understand what tech is better for building your apps! Hope that helps😄

[@hamzatariq8396]

I am not getting results from {movie.$vectorize} instead getting error. Please let me know how to get it. I want to show Plot

[@balaganesh5]