Chat with Your Own Data using Langchain and Streamlit

In this video, we are building a ChatGPT alike chatbot that can answer any questions about your own data.
We continue the series of the previous two videos by, this time, creating it inside your browser using Streamlit!

At the end of the video, you will be able to build your own ChatGPT chatbot that can be run in your browser

⏰ Timestamps:
0:00​ - Demo & Introduction
1:12 - Architecture of the project
2:29 - The view model
4:27 - The UI interface with Streamlit
7:27 - Running the project & Conclusion

🔗 Links

👋🏻 About Me
I am Dona, a software engineer based in Sydney passionated about the wonderful world of AI.

© Copyright resources
Icons:

Audio:

[paullopez_ai]

Dona - this is one of the most comprehensive videos i’ve seen in quite a while! I like the way you have organized the code in the files, you can tell that you know how to build enterprise apps.

[hwlnput]

Good learning materials. Very detailed. Thx!

[nvn]

Thanks for the detailed explanation! Love this series :)

[purple_bread]

Thanks for sharing this! I am learning about chatbots and this is a great tutorial!

[corvuscorax]

Thank you for your guidance, Dona - great tutorial!

[mwsdpbc]

Love this. Maybe you should give classes or run some workshops 🙃

[gplayerone]

Hi! Nice video! thanks for sharing. I just have a question: I'm looking to create an AI agent (chatbot) where I can upload documents (totaling more than 500MB of text storage) and share it as a web application or embed it on a webpage. It should have the following functionalities:

1) The chatbot should allow users to log in using Google Auth, enabling each person to have their own account to use the chat.

2) Monetization capabilities to limit the number of queries to the chat to a maximum of 10 in the free plan. Registered users should be able to subscribe to a paid plan for unlimited access to questions (thus requiring an integrated payment gateway).

3) As the administrator and creator of the chat, I should have access to metrics such as the number of registered users, usage times, most frequent questions or queries, etc.

And think that, also, it should support usage by many people (more than 1000).

How would you recommend carrying out this project? Any specific software, application, or method you would suggest?

[dbarash]

Thanks a lot for this lesson! I have an issue with ingesting, create_chunks() returns 0 even though pdf documents have been uploaded and saved successfully. Any ideas why this exception: langchain\vectorstores\faiss.py", line 562, in __from
index =

IndexError: list index out of range

[zas]

Is there an alternative to using Poetry?