Skip to content

Latest commit

 

History

History
128 lines (90 loc) · 3.17 KB

API_CREDENTIALS_GUIDE.md

File metadata and controls

128 lines (90 loc) · 3.17 KB

API Credentials Setup Guide

🎯 Summary: What You Actually Need

✅ REQUIRED (System won't work without this):

  • Google API Key - For Gemini LLM (the brain of the system)

🔧 OPTIONAL (System works great without these):

  • Google Drive Credentials - Currently using mock responses
  • SerpAPI Key - Currently using free DuckDuckGo

🔑 Required Setup

Google API Key (Gemini LLM) - ESSENTIAL

Why needed: Powers the AI that generates intelligent responses Cost: Free tier available (generous limits)

  1. Go to: https://makersuite.google.com/app/apikey
  2. Click "Create API Key"
  3. Copy the key
  4. Edit backend/.env: 
    GOOGLE_API_KEY=your_actual_key_here

🌐 Optional Enhancements

1. Real Google Drive Access (Optional)

Current Status: ✅ Working with mock responses Why optional: System simulates Google Drive responses for demo

To enable real Google Drive access:

  1. Google Cloud Console: https://console.cloud.google.com/
  2. Create/Select Project
  3. Enable APIs:
    • Go to "APIs & Services" → "Library"
    • Search "Google Drive API" → Enable
  4. Create Credentials:
    • Go to "APIs & Services" → "Credentials"
    • Click "Create Credentials" → "OAuth 2.0 Client IDs"
    • Application type: "Desktop application"
    • Name: "Agentic RAG Chatbot"
  5. Download JSON:
    • Download the credentials file
    • Rename to credentials.json
    • Place in backend/ folder

First run: Browser will open for Google OAuth consent

2. Enhanced Web Search (Optional)

Current Status: ✅ Working with DuckDuckGo (free) Why optional: DuckDuckGo provides good results without API limits

To enable SerpAPI (better results, more reliable):

  1. Sign up: https://serpapi.com/users/sign_up
  2. Get API Key: Dashboard → API Key
  3. Add to .env: 
    SERPAPI_API_KEY=your_serpapi_key_here
  4. Update code in backend/agents/web_search_agent.py: 
    # Uncomment these lines:
from langchain_community.tools import SerpAPIWrapper
search = SerpAPIWrapper(serpapi_api_key=os.getenv("SERPAPI_API_KEY"))

💰 Cost Breakdown

Free Tier Usage:

  • Google Gemini API: 60 queries/minute, 1500/day (very generous)
  • DuckDuckGo: Completely free, unlimited
  • Google Drive API: 100 requests/100 seconds/user (sufficient for most use)

Paid Options:

  • SerpAPI: $50/month for 5000 searches (optional)
  • Google Gemini: $0.001 per 1000 tokens after free tier

🚀 Quick Start Recommendation

For immediate testing: Just add Google API key (only required step) For production: Add all credentials for full functionality

The system is designed to work excellently with just the Google API key!

🔍 How to Check What's Working

Run this command to see current status:

curl http://localhost:8001/health

Response shows which services are active:

{
  "status": "healthy",
  "services": {
    "pdf_processing": "✓",
    "vector_store": "✓",
    "web_search": "✓",
    "google_drive_mcp": "✓",
    "speech_to_text": "✓"
  }
}

All services show ✓ even with mock Google Drive!