🎯 Interview Coach - AI Assistant

Real-time AI-powered interview coaching for Google Meet with local RAG knowledge base.

---

✨ Features

• 🎙️ Live Caption Reading - Reads Google Meet captions in real-time
• 🤖 Multi-AI Provider - OpenAI, Gemini, Groq, OpenRouter, or FREE Pollinations
• 📚 RAG Knowledge Base - Upload your resume, job descriptions, notes
• 📄 File Upload - Supports .txt, .md, .pdf, .docx, .pptx
• 🖱️ Draggable Widget - Float anywhere on screen
• 🔮 Transparent UI - See through to your interview

---

🚀 Quick Start

1. Install Extension

# Clone or download
cd ai-interview-coach

# Load in Chrome
1. Go to chrome://extensions
2. Enable "Developer mode"
3. Click "Load unpacked"
4. Select the ai-interview-coach folder

2. Install PostgreSQL & pgvector (Windows)

1. Download PostgreSQL 18:
   postgresql-18.1-2-windows-x64.exe

2. Download pgvector extension:
   vector.v0.8.1-pg18.zip

   • Extract the zip
   • Copy files to PostgreSQL installation directory:
     • vector.dll → C:\Program Files\PostgreSQL\18\lib\
     • vector.control and vector--*.sql → C:\Program Files\PostgreSQL\18\share\extension\

3. Setup RAG Server

# Install Ollama (for local embeddings)
ollama pull nomic-embed-text

# Create database with pgvector
psql -U postgres -c "CREATE DATABASE salescoach_rag;"
psql -U postgres -d salescoach_rag -c "CREATE EXTENSION vector;"

# Start server
cd server
pip install -r requirements.txt
uvicorn main:app --reload --port 8000

4. Configure

1. Click the extension icon
2. Select AI provider (Pollinations is FREE)
3. Add API key if needed
4. Upload your resume/notes to Knowledge Base

---

📁 Project Structure

cheat-app/
├── manifest.json           # Chrome extension config
├── src/
│   ├── background/
│   │   └── background.js   # AI API calls, RAG integration
│   ├── content/
│   │   ├── sidebar.html    # Floating widget UI
│   │   ├── sidebar.css     # Transparent glass theme
│   │   ├── sidebar.js      # Draggable, widget logic
│   │   ├── meet-captions.js # Caption grabber
│   │   └── content-script.js
│   ├── popup/
│   │   ├── popup.html      # Settings UI
│   │   ├── popup.css       # Glass theme styles
│   │   └── popup.js        # Settings logic, file upload
│   └── utils/
│       ├── constants.js    # Provider configs, RAG settings
│       └── storage.js      # Chrome storage utils
├── server/                 # RAG Backend
│   ├── main.py             # FastAPI endpoints
│   ├── database.py         # PostgreSQL + pgvector
│   └── requirements.txt    # Python dependencies
└── assets/                 # Icons

---

🔧 AI Providers

Provider	API Key	Cost	Models
Pollinations	❌ None	FREE	openai, mistral, llama
OpenAI	✅ sk-...	Paid	gpt-4o-mini, gpt-4o
Gemini	✅ AI...	Free tier	gemini-1.5-flash
Groq	✅ gsk_...	Free tier	llama-3.1-8b
OpenRouter	✅ sk-or-...	Pay-per-use	Multiple

---

📚 RAG Knowledge Base

Upload documents to enhance AI responses with your personal knowledge:

Supported Files

• .txt - Plain text
• .md - Markdown
• .pdf - PDF documents
• .docx - Word documents
• .pptx - PowerPoint presentations

API Endpoints

Endpoint	Method	Description
/health	GET	Server status
/documents	POST	Add document
/documents	GET	List all
/documents/{title}	DELETE	Remove
/search	POST	Vector search
/extract-file	POST	Extract text from file

---

🎨 UI Features

• 🖱️ Draggable - Drag header to move widget
• ➖ Minimize - Collapse to header only
• 📏 Resize - Drag corner to resize
• 💾 Position Memory - Remembers position on reload
• 🔮 Transparent - See-through glass theme

---

💡 Interview Prompt

The AI is configured as an interview candidate:

• Uses RAG knowledge base FIRST
• Simple, natural English
• Short answers (2-3 sentences)
• Honest, no buzzwords
• Practical and logical

---

🛠️ Development

# Watch for changes
# Extension auto-reloads on change

# Server with hot reload
cd server
uvicorn main:app --reload --port 8000

---

📝 License

MIT License - Use freely for your interviews!

---

Made with ❤️ for job seekers