Hospital AI Workflow WhatsApp Bot

I already run a PostgreSQL-backed hospital-management database and now want to layer an intelligent automation flow on top of it. The job breaks down into three closely linked pieces:

1. Connect an AI agent to my existing PostgreSQL instance so it can read/write patient, appointment, billing and inventory tables in real time—safely and with role-based access in mind.
2. Orchestrate an end-to-end workflow that automates the everyday tasks we handle manually today (think appointment confirmations, lab-result notifications, stock re-ordering and discharge summaries). I am happy to discuss the best orchestration stack, but I tend to work in Python and have been looking at LangChain, Zapier Interfaces and similar frameworks.
3. Expose the result through a WhatsApp bot. The bot should understand natural-language requests from staff or patients, call the workflow where appropriate, and answer back with up-to-date information. Twilio’s WhatsApp Business API is already connected; you just need to wire your solution to it.

I will supply database credentials, sample data and an outline of core use-cases. Your deliverable is a working proof-of-concept running on my VPS (Docker preferred) plus clean, well-commented source code. I’ll sign off once:

• The agent performs CRUD operations on PostgreSQL without errors.
• Key hospital tasks run autonomously through the workflow engine.
• The WhatsApp bot returns correct responses for the agreed test scripts.

If you have shipped similar conversational AI integrations—especially with PostgreSQL—please share a short demo or repo link when you bid.