A tool for finding and analyzing local businesses using AI-powered data extraction.
- Node.js 16+
- Ollama (for local LLM)
- SearxNG instance
- Install Ollama:
# On macOS
brew install ollama- Start Ollama:
# Start and enable on login
brew services start ollama
# Or run without auto-start
/usr/local/opt/ollama/bin/ollama serve- Pull the required model:
ollama pull mistral- Clone and set up the project:
git clone https://github.com/yourusername/bizsearch.git
cd bizsearch
npm install- Configure environment:
cp .env.example .env
# Edit .env with your settings- Start the application:
npm run dev- Open http://localhost:3000 in your browser
If Ollama fails to start:
# Stop any existing instance
brew services stop ollama
# Wait a few seconds
sleep 5
# Start again
brew services start ollamaTo verify Ollama is running:
curl http://localhost:11434/api/version- Business search with location filtering
- Contact information extraction
- AI-powered data validation
- Clean, user-friendly interface
- Service health monitoring
Key environment variables:
SEARXNG_URL: Your SearxNG instance URLOLLAMA_URL: Ollama API endpoint (default: http://localhost:11434)SUPABASE_URL: Your Supabase project URLSUPABASE_ANON_KEY: Your Supabase anonymous keyCACHE_DURATION_DAYS: How long to cache results (default: 7)
- Create a new Supabase project
- Run the SQL commands in
db/init.sqlto create the cache table - Copy your project URL and anon key to
.env
MIT
The application uses Supabase for caching search results. Cache entries expire after 7 days.
If automatic cleanup is not available, you can manually clean up expired entries:
- Using the API:
curl -X POST http://localhost:3000/api/cleanup- Using SQL:
select manual_cleanup();View cache statistics using:
select * from cache_stats;