This project is a multi-agent AI system designed to assist patients and dentists throughout the dental treatment journey. The platform combines image analysis, treatment prediction, insurance reasoning, habit coaching, and appointment scheduling into a unified workflow.

Patients begin by uploading images of their teeth. A network of specialized AI agents then analyzes the images, predicts potential treatments, suggests hygiene improvements, evaluates insurance coverage options, and helps schedule dental appointments. All outputs are coordinated by a central Agent Orchestrator, which compiles results into a clear summary for both the patient and the dentist.

The goal of the system is to make dental care more proactive, transparent, and personalized, helping patients understand their treatment path while giving providers better insight before consultations.

The system uses an Agent Orchestrator to coordinate the six main agents in the dental workflow.

flowchart LR
    U[Agent Orchestrator]

    V[Visual Agent]
    F[Financial Agent]
    H[Habit Coaching Agent]
    Sx[Surgery Agent]
    D[Doctor Summary Agent]
    S[Scheduling Agent]

    U --> V
    U --> F
    U --> H
    U --> Sx
    U --> D
    U --> S

Visual Agent: analyzes uploaded teeth images and identifies visible issues.

Financial Agent: retrieves and reasons over Sun Life insurance documentation using RAG.

Habit Coaching Agent: generates personalized hygiene recommendations.

Surgery Agent: runs a scan of the teeth and shows results.

Doctor Summary Agent: orchestrates all outputs into a concise provider report.

Scheduling Agent: pulls dentist information and availability for the user to schedule the next appointment.

1. User Upload
   The patient uploads images of their teeth through the frontend interface.

2. Visual Analysis
   The Visual Agent analyzes the images to identify alignment issues, abnormalities, or potential treatment areas.

3. Treatment Insights
   The Surgery Agent evaluates the scan results and suggests potential procedures or corrective actions.

4. Insurance Evaluation
   The Financial Agent queries Sun Life insurance documentation using Retrieval-Augmented Generation (RAG) to determine coverage options and financial recommendations.

5. Habit Coaching
   The Habit Coaching Agent provides personalized oral hygiene improvements that can help improve treatment outcomes.

6. Doctor Summary Generation
   The Doctor Summary Agent aggregates outputs from all agents and produces a concise diagnostic overview for the dentist.

7. Appointment Scheduling
   The Scheduling Agent retrieves dentist availability and suggests appointment times so the patient can move directly from diagnosis to treatment.

The Agent Orchestrator coordinates these steps and ensures that information flows correctly between agents.

Minimal project scaffold for a patient-first dental workflow with:

• frontend/: Next.js + React + Tailwind shell
• backend/: FastAPI placeholder service

• Frontend: Next.js, React, TailwindCSS
• Backend: Python, FastAPI
• AI / LLM: Google Gemini
• Languages: Python, TypeScript, JavaScript, HTML, CSS
• Architecture: Multi-agent orchestration system with a central Agent Orchestrator
• Retrieval: RAG pipeline with vectorized knowledge base

• User upload
• Shared dental model
• Agent orchestrator
• Treatment predictive agent
• Habit coaching agent
• Financial agent
• Clinic locator and scheduling agent
• Monitoring loop
• Final dashboard

The FastAPI app is organized into folders for the core workflow:

• backend/app/pipeline/dental_model/
• backend/app/core/
• backend/app/agents/treatment_predictive/
• backend/app/agents/habit_coaching/
• backend/app/agents/financial/
• backend/app/agents/clinic_locator/
• backend/app/agents/monitoring/

The scaffold currently exposes:

• GET /
• GET /health
• GET /pipeline/dental-model
• GET /orchestrator
• GET /agents/treatment-predictive
• GET /agents/habit-coaching
• GET /agents/financial
• GET /agents/clinic-locator
• GET /agents/monitoring

From backend/:

python -m venv .venv
source .venv/bin/activate   # (Windows: .venv\Scripts\activate)
pip install -r requirements.txt
uvicorn app.main:app --reload --port 8000

Verify:

• http://localhost:8000/health

From frontend/:

npm install
npm run dev

Open:

• http://localhost:3000