An intelligent, completely local Retrieval-Augmented Generation (RAG) application. Type a question or request a summary of a video, and a local LLM will analyze the transcript text, explain the answer, and seamlessly cut and stitch the corresponding video highlights side-by-side in a web dashboard.

---

🌟 Features

• 🧠 100% Local Intelligence: Powered by Ollama (llama3.2:1b)—no external API keys or cloud dependencies required.
• 🔍 Semantic Video Search: Uses ChromaDB to understand the meaning behind your queries, bypassing simple keyword matching.
• ✂️ Dynamic Slicing & Stitching: Orchestrates MoviePy v2 to instantly crop video frames using precise timestamp structures.
• 📱 Split-Screen UI: Provides a beautiful, parallel Streamlit interface featuring text explanations alongside live video playback.

---

🛠️ The Tech Stack

• Frontend: Streamlit (Wide-mode UI Layout)
• Vector Engine: ChromaDB (Persistent local embeddings storage)
• LLM Orchestrator: Ollama & LangChain
• Video Automation: MoviePy v2 (.subclipped() engine)

---

🚀 Quick Start & Installation

1. Clone the Repository

git clone [https://github.com/Sadhansrid007/V2V_transmission.git]
cd V2V_transmission

2. Install Dependencies

pip install streamlit chromadb langchain ollama moviepy

3. Initialize Your Local Model

ollama pull llama3.2:1b

4. Fire Up the Dashboard!

streamlit run app.py