abhiinnovates/whatsapp-hr-assistant

WhatsApp HR Assistant is a chatbot that lets employees ask HR questions and retrieve personal data through a regular WhatsApp conversation. Employees can ask about leave policies, check how many leave days they have remaining, or request a salary slip PDF, all without logging into a separate HR portal. The assistant connects to WhatsApp via Twilio and uses a fast, free language model from Groq. The assistant handles three types of requests: HR policy questions, live employee data lookups, and salary slip generation. For policy questions, it searches a vector database built from your uploaded HR documents and returns an answer with source references, so it only tells employees what your documents actually say. For personal data like leave balances and salary slips, it calls a connected HR system, then generates and sends the response back on WhatsApp. Before sharing any personal information, the assistant verifies identity using a one-time code. When an employee asks for their leave balance or salary slip, the bot sends a six-digit code to their registered work email. Once the employee types the code back in the WhatsApp chat, the original question is answered automatically. A verified session lasts 30 minutes so employees are not re-prompted for every message in that window. The HR system integration uses a mock module by default, which makes it possible to run the whole project locally without a real HR backend. The mock exposes a small interface for employee lookup, leave balances, and salary structure. The README explains how to replace those functions with calls to a real service such as Zoho People. Setup requires API keys from Groq and Twilio, a public URL (such as an ngrok tunnel) so Twilio can reach the local server, and a one-time step to ingest your HR documents into the FAISS vector store. The project includes test scripts to simulate WhatsApp messages locally without actually sending them.