How haswinai/recommendation-and-negotiation works
Main Components
A. Frontend Layer:
• Web Interface (HTML/CSS/JavaScript)
• User Interaction Components:
• Analysis Section
• Negotiation Interface
• Post-Call Summary View
B. Backend Layer (Flask Application):
• Main Application Server (app.py)
• Integration Services:
• GROQ API Integration (groq_integration2.py)
• Google Sheets Integration (google_sheets_util.py)
• Data Processing Components:
• Text Recognition System
• Mutual Funds Dataset Handler
C. External Services:
• GROQ API (LLM Service)
• Google Sheets API (Data Storage)
• OAuth2 Authentication Service
[Client Tier] → [Application Tier] → [External Services Tier]
Component Relationships to Highlight
A. Data Flow Paths:
• User Interface → Flask Server → GROQ API
• → Google Sheets
B. Authentication Flow:
• App → OAuth2 → Google Services
C. Processing Pipeline:
• Input → Text Recognition → LLM Processing → Recommendations
Technical Details to Note
• HTTP/HTTPS endpoints
• API integration points
• Data persistence locations
• Authentication checkpoints
• Load balancing points (if any)
Specific Implementation Notes
• Real-time processing capabilities
• Integration with LLM services
• Data storage mechanisms
• Security implementation
• User interaction flows
Remember to:
• Use consistent styling
• Version information
• Notes for critical components
• System boundaries