BizLaunch Navigator

An AI-powered web application that helps entrepreneurs and non-profit founders get started by intelligently connecting them with mentors, business registration resources, suppliers, grants, and safety information.

Features

• 🤖 LLM-Powered: Uses Google Gemini to understand natural language queries and extract relevant information
• 📚 SCORE Mentors: Find industry and expertise-matched mentor leads
• 📄 Business Registration: Get California business registration forms, PDFs, and fees
• 🏭 Supplier Search: Find North American suppliers for raw materials and products
• 💰 Grant Opportunities: Discover funding opportunities from the California Grant Portal
• 🛡️ Safety Assessment: View neighborhood crime data for business location planning
• 🎨 Neubrutalism UI: Bold, vibrant, high-contrast design

Tech Stack

Backend

• FastAPI: Modern Python web framework
• Google Gemini: Free LLM for parameter extraction
• Browser Use SDK: Integration with 5 custom skills
• Pydantic: Data validation and settings management

Frontend

• React: UI library
• Vite: Build tool and dev server
• Axios: HTTP client
• Neubrutalism CSS: Bold, vibrant design system

Setup Instructions

Prerequisites

• Python 3.8+
• Node.js 16+
• Browser Use API key
• Google Gemini API key (free tier)

Backend Setup

1. Navigate to the backend directory:

cd backend

2. Create a virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

pip install -r requirements.txt

4. Create a .env file in the backend directory:

BROWSER_USE_API_KEY=your_browser_use_api_key_here
GEMINI_API_KEY=your_gemini_api_key_here
FRONTEND_URL=http://localhost:5173

5. Run the backend server:

uvicorn app.main:app --reload --port 8001

The API will be available at http://localhost:8001

Frontend Setup

1. Navigate to the frontend directory:

cd frontend

2. Install dependencies:

npm install

3. Start the development server:

npm run dev

The app will be available at http://localhost:5173

Usage

1. Start both the backend and frontend servers
2. Open http://localhost:5173 in your browser
3. Type a natural language query about your business idea, for example:
   • "I want to start a coffee shop in San Francisco"
   • "Find me a mentor for tech startups"
   • "I need suppliers for biodegradable packaging in Los Angeles"
   • "What grants are available for small businesses in California?"

The AI will:

• Extract relevant information from your query
• Determine which skills to call
• Execute the appropriate API calls
• Present formatted results

Project Structure

small-business-orchestrator/
├── backend/
│   ├── app/
│   │   ├── __init__.py
│   │   ├── main.py              # FastAPI app entry point
│   │   ├── models.py            # Pydantic models
│   │   ├── llm_service.py       # Gemini LLM integration
│   │   ├── skill_orchestrator.py # Browser Use skill execution
│   │   └── config.py            # Configuration
│   ├── requirements.txt
│   └── .env.example
├── frontend/
│   ├── src/
│   │   ├── App.jsx
│   │   ├── components/
│   │   │   ├── ChatInterface.jsx
│   │   │   ├── ResultsDisplay.jsx
│   │   │   └── SkillCard.jsx
│   │   ├── styles/
│   │   │   └── neubrutalism.css
│   │   └── index.js
│   ├── package.json
│   └── vite.config.js
└── README.md

API Endpoints

POST /api/chat

Main chat endpoint that processes user queries.

Request:

{
  "message": "I want to start a coffee shop in San Francisco",
  "conversation_history": [] // optional
}

Response:

{
  "message": "Found 5 mentor(s). Found 3 supplier(s)...",
  "skills_executed": ["score_mentor", "thomasnet", "safety"],
  "results": {
    "score_mentor": {
      "skill_name": "score_mentor",
      "success": true,
      "data": { ... }
    }
  },
  "follow_up_questions": []
}

Environment Variables

• BROWSER_USE_API_KEY: Your Browser Use API key (required)
• GEMINI_API_KEY: Your Google Gemini API key (required)
• FRONTEND_URL: Frontend URL for CORS (default: http://localhost:5173)

Notes

• The Thomasnet skill requires a datadome cookie. The Browser Use SDK should handle this automatically.
• All skills are called in parallel for better performance.
• The LLM intelligently determines which skills to call based on user input.
• If critical information is missing, the system will ask follow-up questions.

License

MIT