This is a Simple Chat Application that allows users to search youtube videos and give the URL to the chat app and then chat with AI ChatBot regarding the content of the video All-in-one-place.
I implemented this using OpenAI Whisper, OpenAI GPT-3.5 Turbo and Text Embedding Ada-002 to enhance the model's performance. Utilizing Pinecone Vector database to store documents for fast query and retrieval.
It works like:
Youtube URL -> OpenAI Whisper (transcription) -> (text splitter , embeddings) Pinecone Documents
User Query -> Pinecone -> GPT-3.5 -> User
User can keep on adding videos data while searching for certain topic or choose to clear the Vector DB and start fresh!
Also i have implemented a simple chatgpt below for any video unrelated queries or tasks the user might have.
Gurjit11
commented
9 months ago
Project name
VideoQuery.ai
Description
This is a Simple Chat Application that allows users to search youtube videos and give the URL to the chat app and then chat with AI ChatBot regarding the content of the video All-in-one-place.
I implemented this using OpenAI Whisper, OpenAI GPT-3.5 Turbo and Text Embedding Ada-002 to enhance the model's performance. Utilizing Pinecone Vector database to store documents for fast query and retrieval.
It works like: Youtube URL -> OpenAI Whisper (transcription) -> (text splitter , embeddings) Pinecone Documents
User Query -> Pinecone -> GPT-3.5 -> User
User can keep on adding videos data while searching for certain topic or choose to clear the Vector DB and start fresh! Also i have implemented a simple chatgpt below for any video unrelated queries or tasks the user might have.
Language
English
Project Repository URL
https://github.com/Gurjit11/AI-Chat-App-Hack
Deployed Endpoint URL
https://video-query-eight.vercel.app/
Project video
https://www.loom.com/share/5334ba72f1ba4eeb8b438d5a8b4d671b?sid=7b68cfb2-43fd-467b-b8c0-bb7359e70572
Team members
Gurjit11
Showcase Consent
Yes