🚀 AI-Powered Chrome Extensions Portfolio

A Collection of Intelligent Browser Extensions Built with LLMs, RAG, and Modern Web APIs

Three production-ready Chrome extensions showcasing AI/ML integration — from contract analysis to video Q&A to voice-powered assistants.

---

📦 Projects

This repository contains three standalone AI-powered Chrome extensions, each with its own FastAPI backend and dedicated README.

#	Project	Description	LLM	Key Tech
1	⚖️ Legal Agreement Analyzer	Upload contracts & get AI risk analysis, clause breakdowns, and legal Q&A	Gemini 2.5 Flash	RAG, ChromaDB, HuggingFace, PyPDF2
2	🎙️ Text-to-Voice LLM Extension	Select text on any page → local AI explains it → speaks response aloud	Llama 3 (Ollama)	Web Speech API, Ollama
3	🎬 YouTube RAG Assistant	Ask questions about any YouTube video using its transcript	Gemini 2.5 Flash	RAG, ChromaDB, YoutubeLoader

---

⚖️ Legal Agreement Analyzer

AI-powered contract risk analysis as a Chrome extension

📖 Click to expand details

What It Does

• Upload PDF/TXT legal documents or analyze any webpage
• Get a risk score (0–100%) with severity-tagged key risk points
• Ask questions in natural language and get clause-level answers
• Premium UI with gold/dark theme, glassmorphism, and animations

Tech Stack

FastAPI · LangChain · Gemini 2.5 Flash · ChromaDB · HuggingFace Embeddings · PyPDF2 · Chrome Extension MV3

Quick Start

cd legal_agreement_analyzer/backend
pip install -r requirements.txt
# Add GOOGLE_API_KEY to .env
uvicorn api:app --reload --port 8000

Then load the extension/ folder in chrome://extensions/.

→ Full README

---

🎙️ Text-to-Voice LLM Extension

Select text → AI explains it → speaks it aloud — all locally

📖 Click to expand details

What It Does

• Select any text on a webpage
• Sends it to a local Llama 3 model (via Ollama) — no cloud needed
• Displays the AI response in a neon-themed popup
• Speaks the response using the Web Speech API

Tech Stack

FastAPI · Ollama · Llama 3 · Web Speech API · TailwindCSS · Chrome Extension MV3

Quick Start

ollama pull llama3
cd text-to-voice-llm-extension
pip install fastapi uvicorn ollama
uvicorn app:app --reload

Then load the frontend/ folder in chrome://extensions/.

→ Full README

---

🎬 YouTube RAG Assistant

Chat with any YouTube video using its transcript

📖 Click to expand details

What It Does

• Enter a YouTube Video ID
• Ask any question about the video
• Get transcript-grounded answers powered by a full RAG pipeline
• Dark, indigo-themed chat interface with animated message bubbles

Tech Stack

FastAPI · LangChain · Gemini 2.5 Flash · ChromaDB · HuggingFace Embeddings · YoutubeLoader · Chrome Extension MV3

Quick Start

cd yt_video_chatbot/backend
pip install fastapi uvicorn python-dotenv langchain langchain-google-genai langchain-community chromadb sentence-transformers
# Add GOOGLE_API_KEY to .env
uvicorn api:app --reload --port 8000

Then load the frontend/ folder in chrome://extensions/.

→ Full README

---

🧠 Shared Concepts & Skills

All three projects demonstrate proficiency in:

Skill	Description
🤖 LLM Integration	Working with both cloud (Gemini) and local (Llama 3) language models
📚 RAG Pipelines	Retrieval-Augmented Generation — embedding, vector search, context-grounded answers
🌐 Chrome Extension Dev	Manifest V3, popup UIs, chrome.scripting, chrome.storage
⚡ FastAPI Backends	Async Python REST APIs with CORS, file uploads, and structured responses
🗄️ Vector Databases	ChromaDB for semantic search over document/transcript chunks
🎨 Modern UI/UX	Glassmorphism, gradients, animations, dark themes, responsive design
🔒 Privacy-First Design	In-memory processing, local LLM support, no persistent storage

---

📁 Repository Structure

own_project/
│
├── legal_agreement_analyzer/     ⚖️  Contract risk analysis extension
│   ├── backend/                      FastAPI + LangChain RAG backend
│   ├── extension/                    Chrome Extension (MV3)
│   ├── sample_agreement.txt          Test legal document
│   └── README.md                     Full documentation
│
├── text-to-voice-llm-extension/  🎙️  Local LLM + voice extension
│   ├── app.py                        FastAPI + Ollama backend
│   ├── frontend/                     Chrome Extension (MV3)
│   └── README.md                     Full documentation
│
├── yt_video_chatbot/             🎬  YouTube transcript Q&A extension
│   ├── backend/                      FastAPI + LangChain RAG backend
│   ├── frontend/                     Chrome Extension (MV3)
│   └── README.md                     Full documentation
│
└── README.md                     📄  ← You are here (this file)

---

🚀 Getting Started (All Projects)

Common Prerequisites

• Python 3.8+ → python.org
• Google Chrome → Latest version
• Google API Key (for Gemini projects) → Get one here
• Ollama (for Text-to-Voice project) → ollama.com

General Workflow

1. Navigate into the project folder
2. Install Python dependencies (pip install -r requirements.txt or manually)
3. Configure environment variables (.env with API keys)
4. Start the FastAPI backend (uvicorn ... --reload)
5. Load the Chrome extension via chrome://extensions/ → Developer mode → Load unpacked
6. Use the extension from the Chrome toolbar!

---

📄 License

All projects are licensed under the MIT License — feel free to study, modify, and build upon them.

---

Crafted with ❤️ — AI Extensions for the Real World

From contract analysis to video Q&A to voice assistants — powered by cutting-edge AI.