Predict how humans respond to health interventions before running real-world trials.
Built for AI researchers, digital health builders, and multi-agent experimentation.
HealthAgent is a multi-agent simulation framework that models how human emotion, behavior, and physiology interact during health interventions.
Instead of running expensive real-world trials, HealthAgent allows researchers and builders to:
✔ simulate long-term patient trajectories
✔ test behavioral interventions
✔ evaluate compliance strategies
✔ study digital health systems
All inside a fast AI-powered simulation environment.
Think of it as a health behavior simulator for digital health research.
Testing health interventions in the real world is:
- expensive
- slow
- ethically constrained
You cannot easily run large randomized trials just to test a new behavioral nudge.
HealthAgent allows you to simulate those trials in seconds.
The system models realistic human behavior patterns:
- compliance rates around 40–60% for high-stress individuals
- emotional fluctuation driven by sleep and fatigue
- physiological response to behavioral adherence
Example output from a 90-day simulation.
HealthAgent models health behavior using collaborating agents.
Orchestrator Agent
├ Emotion Agent
│ ├ stress
│ ├ motivation
│ └ fatigue
│
├ Compliance Agent
│ ├ decision probability
│ └ adherence score
│
├ Physiology Agent
│ ├ HRV
│ ├ resting HR
│ └ sleep dynamics
│
└ Intervention Agent
├ risk detection
└ behavioral nudges
Simulation loop:
Emotion → Compliance → Physiology → Intervention → Record
Clone repository
git clone https://github.com/csong8904-spec/ReHealthAI-HealthAgent.git
cd ReHealthAI-HealthAgent
Create environment
python -m venv venv
source venv/bin/activate
Install dependencies
pip install -r requirements.txt
Add API key
DEEPSEEK_API_KEY=your_key_here
Run simulation
python simulate.py
Visualize results
python visualize.py
from healthagent import SimulationEngine
from healthagent.models import PatientProfile, InterventionPlan
patient = PatientProfile(
age=52,
bmi=27.5,
stress_level="HIGH",
sleep_avg_hours=6.2
)
plan = InterventionPlan(
name="Lifestyle Optimization",
duration_days=90,
rules=[
"30-minute walk daily",
"sleep before 10:30 PM",
"10 min mindfulness"
]
)
engine = SimulationEngine(seed=42)
result = engine.run_simulation(patient, plan)
print(result.compliance_rate)
print(result.sleep_improvement)
print(result.hrv_change)
ReHealthAI-HealthAgent
│
├ healthagent
│ ├ agents
│ │ ├ emotion_agent.py
│ │ ├ compliance_agent.py
│ │ ├ physiology_agent.py
│ │ ├ intervention_agent.py
│ │ └ orchestrator.py
│ │
│ ├ models
│ │ ├ patient_profile.py
│ │ ├ intervention.py
│ │ └ trajectory.py
│ │
│ └ engine
│ └ simulation_engine.py
│
├ examples
│ ├ sleep_simulation.py
│ └ cardio_simulation.py
│
├ simulate.py
├ visualize.py
├ requirements.txt
Digital Health Research
Study behavioral drivers behind intervention adherence.
Healthcare Analytics
Estimate success probability of interventions.
Digital Therapeutics
Prototype behavioral health treatments.
AI Agent Research
Experiment with LLM-driven agent simulations.
HealthAgent is part of the ReHealth AI platform.
Our long-term vision is to build a digital twin simulation environment for human health.
Fork repository
git checkout -b feature/my-feature
Commit changes
git commit -m "Add feature"
Push branch
git push origin feature/my-feature
Open Pull Request.
Apache 2.0 License
If this project is useful:
Star the repo
Fork it
Report issues
Share it
Built with ❤️ by ReHealth AI · 2026