# Entity Comparison Tool

This project demonstrates how to integrate the Qdrant vector database with the Tavily search API to generate a markdown table comparing three entities. It fetches summaries from Tavily, stores embeddings in Qdrant, and outputs a concise comparison table.

## Features

- **Qdrant Integration**: Stores and retrieves vector embeddings for entities.
- **Tavily Search**: Retrieves up-to-date summaries and URLs for each entity.
- **Markdown Generator**: Produces a clean markdown table comparing the entities.

## Prerequisites

- Python 3.9+
- A running Qdrant instance (default: `localhost:6333`)
- A Tavily API key

## Setup

```bash
# Clone the repository
git clone https://github.com/yourusername/entity-comparison-tool.git
cd entity-comparison-tool

# Create a virtual environment
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt
```

## Configuration

Create a `.env` file in the project root with the following variables:

```dotenv
# Tavily API key
TAVILY_API_KEY=your_tavily_api_key

# Qdrant connection (optional, defaults to localhost:6333)
QDRANT_HOST=localhost
QDRANT_PORT=6333
QDRANT_COLLECTION=entities
```

## Usage

```bash
# Compare three entities
python src/main.py "Python" "Java" "C++"
```

The script will output a markdown table similar to:

```markdown
| Attribute | Python | Java | C++ |
|-----------|--------|------|-----|
| Summary | Python is a high-level, interpreted programming language... | Java is a class-based, object-oriented programming language... | C++ is a general-purpose programming language that supports procedural, object-oriented, and generic programming... |
| URL | https://www.python.org/ | https://www.oracle.com/java/ | https://isocpp.org/ |
```

## Project Structure

```
entity-comparison-tool/
├── src/
│   ├── main.py
│   ├── qdrant_client.py
│   └── markdown_generator.py
├── requirements.txt
└── README.md
```

## Extending the Tool

- **Custom Attributes**: Modify `markdown_generator.py` to extract additional attributes from the Tavily response.
- **Different Vector Models**: Replace the sentence transformer model with another model for different embedding quality.
- **Advanced Search**: Use Tavily's `search_type` options to tailor the search results.

## License

MIT License