Chat With Your Documents Using LangChain + JavaScript

In this session we will go over how to build a a chatbot similar to ChatGPT that can answer questions about your specific data.

Will be built end-to-end with LangChain's Javascript package

[leonvanzyl]

I've been creating Langchain tutorials for a while now, and never once thought about looking up your official channel. Really glad I found this 😂.
Keep it up guys!

[Camxlare]

Holy Sh*^\ this video is so underrated. @37:30, I was hoping so bad you guys was a bit patient to let it run but it’s all good! I love what you guys are doing. I’ll be definitely using this and supporting this project.

[SLTRM]

Any example how load multiple docs to the same vector store and ask over info of all documents at same time? For example a book or report with multiple volumes (files). Where the volume/file number is important. "In what volume talk about of recesion"

[stalingagnay]

great video !
What tool did you use to generate those embeddigns graphics?

[htmlfivedev]

The audio volume jumps up and down like crazy ... impossible to follow ...

[JohnnyLonghorn]

didn't catch the link that was mentioned @50:28

[joshuashepherd]

36:05 Poor bro was trying to run a .ts export module as an npm script 😵
I feel for you tho. Typescript has made my dev life like 30% better, but 80% more frustrating.

[sunilvurandur]

Hey greate work on this. I wonder how do you save the vectorstore into database. Forming an embeddings, vectostore from file is an expensive process. once this is done i can store this vectorstore in to my database and use this often when ever user gives a questions request. I was working on this for days. can you help me to store this vectorstores into database by compressing or something. Thanks in advance

[ejeingu]

why they don't prepare the code before the meeting????!!