Project

OmniBrief

Built a full-stack video intelligence engine that ingests YouTube videos and local files, generating transcripts, summaries, decisions, and action items through AI-powered analysis. Additionally, implemented a RAG-based conversational interface using ChromaDB and LangChain, enabling users to query video content with contextual accuracy.

OmniBrief screenshot

Tech Stack

  • Python
  • ChromaDB
  • Flask
  • Mistral AI
  • LangChain
  • OpenAI Whisper
  • JavaScript
  • React

OmniBrief is a full-stack video intelligence engine that ingests YouTube videos and local files, generating transcripts, summaries, decisions, and action items through AI-powered analysis. Additionally, implemented a RAG-based conversational interface using ChromaDB and LangChain, enabling users to query video content with contextual accuracy.

What it does

  • Video ingestion and processing
  • AI-powered transcript generation
  • Summary and action item extraction
  • RAG-based conversational interface

Key Features

  • Multi-source Ingestion — Supports both YouTube URLs and local video file uploads
  • AI Analysis — Uses Mistral AI for generating transcripts, summaries, decisions, and action items
  • RAG Interface — ChromaDB + LangChain for context-aware video content querying
  • Full-stack Implementation — React frontend with Flask backend API

Tech Stack

Python · Flask · Mistral AI · ChromaDB · LangChain · OpenAI Whisper · JavaScript · React

Command Palette

Search for a command to run...