HealthSense AI — Multi-Agent Healthcare System

A 5-agent orchestration system built with n8n that routes a patient's symptom description through diagnosis, doctor matching, hospital comparison, and appointment booking — end-to-end without human intervention.

Why I Built This

Healthcare navigation is genuinely difficult — most people don't know which specialist they need, which hospitals in their area handle their condition, or how to get an appointment quickly. This project started as a question: how many of those steps could an AI agent handle autonomously? The answer turned out to be all of them. HealthSense AI demonstrates that multi-agent coordination — where each agent is scoped to one job and hands off to the next — can handle complex, multi-step workflows that normally require multiple human interactions across different systems.

Agent Architecture

Patient Input (symptoms)
       │
       ▼
┌─────────────────────┐
│  Coordinator Agent  │  — routes input to the right specialist agent
└─────────────────────┘
       │
   ┌───┴───────────────────────────────────┐
   ▼                                       ▼
Diagnostics Agent                 Hospital Comparison Agent
(symptom → likely condition)      (rank hospitals by specialty + distance)
   │                                       │
   ▼                                       ▼
Doctor Booking Agent              Booking Update Agent
(find available doctor + slot)    (confirm & notify patient)

Agents

Agent	Workflow File	Role
Coordinator	Coordinator_Agent.json	Parses input, routes to specialist agents
Diagnostics	Diagnostics_Agent.json	Maps symptoms → probable diagnosis
Doctor Booking	Doctor_Booking_Agent.json	Finds available slot + matches specialist
Hospital Comparison	Hospital_Comparison_Agent.json	Ranks hospitals by fit
Booking Update	Booking_Agent_Update.json	Confirms appointment, updates records

Data

File	Description
doctors_info.xlsx	Doctor profiles (specialty, hospital, availability)
doctors_slots.xlsx	Available appointment slots
hospitals_info.xlsx	Hospital directory with specialties
hospitals_emergency.xlsx	Emergency-capable facilities
lab_tests.xlsx	Lab test catalog for diagnostic reference

Tech Stack

• Workflow Orchestrator: n8n
• LLM: OpenAI GPT-4o
• Data: Excel/XLSX reference data (doctors, hospitals, slots)
• Pattern: Multi-agent with coordinator routing

Setup

1. Import all .json workflow files into n8n
2. Upload data files to n8n as accessible references
3. Configure OPENAI_API_KEY in n8n credentials
4. Start with Coordinator_Agent.json as the entry point

Use Case

Built to demonstrate agentic AI applied to healthcare operations — showing how LLM reasoning + structured data + multi-agent coordination can automate a complex, real-world workflow that typically requires multiple human touchpoints.