SQL-Agent

Overview

SQL-Agent is a Python-based tool designed to perform similarity searches on a database of documents containing frequently asked questions (FAQs) or similar content. By retrieving relevant documents, SQL-Agent reduces hallucinations and improves the accuracy of AI-generated responses. The tool is ideal for enhancing the reliability of language model outputs by providing contextually relevant information from trusted sources.

---

Features

• Similarity Search: Efficiently searches for documents with questions similar to the input query.
• Reduced Hallucinations: Minimizes errors in AI-generated responses by providing factual references.
• Seamless Database Integration: Connects to PostgreSQL databases for storing and retrieving documents.
• Customizable: Easily adaptable to different datasets and configurations.

---

Requirements

The following Python libraries are required to run SQL-Agent:

• python-dotenv: For managing environment variables.
• langchain_openai: Integration with OpenAI's language models.
• langchain_community: Community-driven tools for LangChain.
• langchain_core: Core LangChain functionalities.
• langchain_experimental: Experimental features for LangChain.
• psycopg2-binary: PostgreSQL database adapter for Python.
• google_search_results: API for Google Search integration.

To install the dependencies, run:

pip install -r requirements.txt

---

Installation

1. Clone the repository:

   git clone https://github.com/PragalvhaSharma/SQL-Agent.git

2. Navigate to the project directory:

   cd SQL-Agent

3. Create a virtual environment (optional but recommended):

   python -m venv venv
   source venv/bin/activate  # On Windows, use `venv\Scripts\activate`

4. Install dependencies:

   pip install -r requirements.txt

---

Usage

1. Environment Setup: Create a .env file in the project directory with the following variables:

   OPENAI_API_KEY=your_openai_api_key
   DATABASE_URL=your_postgresql_database_url
   GOOGLE_API_KEY=your_google_api_key
   GOOGLE_CX=your_google_custom_search_engine_id

2. Database Configuration: Ensure your PostgreSQL database is set up and populated with the required documents. Use the provided SQL schema (if available) to structure your database correctly.

3. Running the Script: Execute the main Python script to start the similarity search:

   python main.py

4. Query Example: Input a question or query, and the system will retrieve similar documents from the database to enhance the response quality.

---

Contributing

Contributions are welcome! If you have ideas, suggestions, or bug fixes, please open an issue or submit a pull request.

---

License

This project is licensed under the MIT License. See the LICENSE file for details.

---

Contact

For questions or support, contact Pragalvha Sharma.

---

Acknowledgments

Special thanks to the LangChain community and contributors for providing tools that made this project possible.