This project provides a prototype autonomous HR chatbot designed to answer employee queries using a suite of tools. It targets HR departments and employees seeking quick, automated access to HR information, leveraging LLM capabilities for efficient query resolution.

How It Works

The chatbot utilizes LangChain's agent and tools framework, powered by ChatGPT (or gpt-3.5-turbo). It accesses information through three primary tools: a Pinecone index containing embeddings of HR policies, a Pandas DataFrame of dummy employee data manipulated via PythonAstREPLTool, and LangChain's built-in calculator for mathematical operations. This multi-tool approach allows the agent to dynamically select and use the most appropriate resource to answer user questions.

Quick Start & Requirements

• Install dependencies: pip install -r requirements.txt
• Requires Python 3.10.
• API keys for OpenAI and Pinecone must be configured.
• Run the application: streamlit run hr_agent_frontent.py
• Embeddings creation requires running store_embeddings_in_pinecone.ipynb.
• Official Demo: Youtube Link

Highlighted Details

• Integrates ChatGPT/GPT-3.5-turbo with LangChain agents and tools.
• Uses Pinecone for vector storage of HR policies.
• Leverages Pandas and PythonAstREPLTool for structured employee data.
• Built with a Streamlit frontend for easy deployment.

Maintenance & Community

• Developed by Stephen Bonifacio.
• Contact: LinkedIn, Twitter

Licensing & Compatibility

• The repository does not explicitly state a license.

Limitations & Caveats

The project is described as a prototype enterprise application. The employee data is dummy data, and the HR policy document is ChatGPT-generated, indicating potential limitations in real-world accuracy and completeness.