Deep Research Assistant

Current Date and Time (UTC): 2025-02-28 13:45:49
Current User: srikrishnavansi
Version: 1.0.0

System Architecture

High-Level Overview

%%{init: {'theme': 'dark', 'themeVariables': { 'fontSize': '16px'}}}%%
flowchart TD
    subgraph FE[Frontend Layer]
        UI[User Interface] --> SM[Session Manager]
    end
    
    subgraph AL[Application Layer]
        RC[Research Controller] --> SC[Synthesis Controller]
    end
    
    subgraph IL[Integration Layer]
        TS[Tavily Search] --> GM[Gemini Model]
    end

    SM --> RC
    SC --> TS
    SC --> GM

Loading

Core Components

1. Research System Flow

%%{init: {'theme': 'dark', 'themeVariables': { 'fontSize': '14px'}}}%%
sequenceDiagram
    participant U as User
    participant R as Research Controller
    participant T as Tavily API
    participant G as Gemini API
    
    U->>R: Submit Query
    R->>T: Execute Search
    T-->>R: Raw Results
    R->>G: Process & Synthesize
    G-->>R: Final Results
    R->>U: Return Response

Loading

2. Data Processing Pipeline

%%{init: {'theme': 'dark', 'themeVariables': { 'fontSize': '14px'}}}%%
graph LR
    A[Query Input] --> B[Validation]
    B --> C[Research]
    C --> D[Synthesis]
    D --> E[Results]

Loading

Core Features

1. Research Capabilities

• Basic Research Mode

  • Quick surface-level analysis
  • Key points extraction
  • Summary generation

• Advanced Research Mode

  • In-depth analysis
  • Comprehensive data gathering
  • Detailed synthesis

2. Data Processing

• Query validation and preprocessing
• Multi-source research integration
• AI-powered synthesis
• Result caching and optimization

3. Result Management

• Structured data format
• History tracking
• Cache management
• Error handling

Technical Implementation

1. Core Configuration

class Settings(BaseSettings):
    # API Keys
    tavily_api_key: SecretStr
    google_api_key: SecretStr
    
    # Research Settings
    max_results_per_query: int = 10
    search_depth: SearchDepth = SearchDepth.BASIC
    max_retries: int = 3
    
    # Session Info
    current_timestamp: datetime = "2025-02-28 13:45:49"
    current_user: str = "srikrishnavansi"

2. Error Handling Strategy

%%{init: {'theme': 'dark', 'themeVariables': { 'fontSize': '14px'}}}%%
graph TD
    A[Error Detection] --> B{Error Type}
    B -->|API| C[Retry Logic]
    B -->|Validation| D[User Feedback]
    B -->|System| E[Logging]
    C --> F[Resolution]
    D --> F
    E --> F

Loading

System Requirements

Software

• Python 3.8+
• Required APIs:
  • Tavily Search API
  • Google Gemini API

Hardware

• RAM: 4GB minimum
• Storage: 1GB minimum
• Stable internet connection

Dependencies

streamlit>=1.24.0
pydantic>=2.0.0
requests>=2.31.0
python-dotenv>=1.0.0

Development Roadmap

Phase 1 (Completed)

✅ Core research functionality ✅ API integrations ✅ Basic synthesis ✅ Error handling

Phase 2 (Current)

🔄 Advanced caching 🔄 Multiple research strategies 🔄 Performance optimization

Phase 3 (Planned)

📅 Machine learning integration 📅 Custom research templates 📅 Advanced analytics

Project Structure

deep_research/
├── app.py                  # Main application
├── config.py              # Configuration
├── static/                # Static assets
├── agents/                # Core agents
│   ├── research_agent.py
│   └── synthesis_agent.py
├── models/                # Data models
├── utils/                 # Utilities
├── requirements.txt
└── README.md

Getting Started

1. Clone the repository
2. Create .env file:

TAVILY_API_KEY=your-tavily-api-key
GOOGLE_API_KEY=your-google-api-key

3. Install dependencies:

pip install -r requirements.txt

4. Run the application:

streamlit run app.py

Documentation

Detailed documentation available for:

• API Integration
• Research Strategies
• Data Models
• Error Handling
• Cache Management
• Results Processing

Contributing

Please refer to CONTRIBUTING.md for:

• Code standards
• Pull request process
• Development guidelines