All purpose RAG chat API (and single page application)
Description
Why enter
I entered the RAG hackaton as a fun challenge to try and learn new things (with a way too tight deadline 😬). While I've worked with AI assisted applications in previous Hack Together project, I had yet to work with Retrieval Augmented Generation. With the goal of learning new things I've not only looked solely at creating a RAG based app, but also using other technologies I haven't had a chance to use before.
Being a hackaton there's still some rough edges and plenty of room for improvements, but ultimately the application is fully functional and I'm proud at how much I've been able to setup in the short amount of time 😀.
Learning Objectives
Creating a RAG application
Setting up Github Actions workflows (with other pipeline experience)
Using vector databases
Working with the OpenAi assistant feature
Whom it's made for
The main feature of the app is its API creating as a cost effective fully Azure cloud-based application. With its easy to use API operations it's capable of allowing developer throughout an organisation to quickly implement RAG-based chat features using their own knowledge base without requiring deeper knowledge of its implementation.
How did RAG help me
Shoutout to Github Copilot for helping me with the streamed responses!
The demo version of the application as seen in the video. Please use responsibly and don't overload it. There's a rate limited that could cause some errors.
Project Name
All purpose RAG chat API (and single page application)
Description
Why enter
I entered the RAG hackaton as a fun challenge to try and learn new things (with a way too tight deadline 😬). While I've worked with AI assisted applications in previous Hack Together project, I had yet to work with Retrieval Augmented Generation. With the goal of learning new things I've not only looked solely at creating a RAG based app, but also using other technologies I haven't had a chance to use before.
Being a hackaton there's still some rough edges and plenty of room for improvements, but ultimately the application is fully functional and I'm proud at how much I've been able to setup in the short amount of time 😀.
Learning Objectives
Whom it's made for
The main feature of the app is its API creating as a cost effective fully Azure cloud-based application. With its easy to use API operations it's capable of allowing developer throughout an organisation to quickly implement RAG-based chat features using their own knowledge base without requiring deeper knowledge of its implementation.
How did RAG help me
Shoutout to Github Copilot for helping me with the streamed responses!
Technology & Languages
Project Repository URL
https://github.com/marcel-paalvast/rag-chat
Deployed Endpoint URL
https://ragchatmeb7a2fcvura4.z6.web.core.windows.net/
The demo version of the application as seen in the video. Please use responsibly and don't overload it. There's a rate limited that could cause some errors.
Project Video
https://github.com/user-attachments/assets/099015cc-6bef-4ad8-9b0f-3ae5e7e93903
Team Members
marcel-paalvast