SeeMyBodyAI Backend

A backend platform built for AI-driven clinic workflows, with a strong focus on RAG based chatbot agent interaction, image generation/editing, and a robust production-ready API.

What this project does

SeeMyBodyAI Backend powers the AI and ML experience for medical aesthetics clinics by combining:

• a chatbot/RAG agent for clinic assistant interaction and guided conversational workflows
• image generation using Google Gemini to simulate patient transformation results
• image editing workflows for refined simulated outputs
• secure admin and user authentication with JWT tokens
• lead capture and analytics through simulation details and consultation records
• payment orchestration using Stripe checkout

Why this project is useful

AI & ML-first capabilities

• Chatbot assistant: provides an authenticated /user/agent endpoint for natural language, clinic-specific responses.
• Generative image pipeline: supports uploading real patient photos, generating a simulation, and editing the output.
• Prompt-driven experience: integrates prompt logic that is ready for extension and experimentation.

Strong backend foundation

• FastAPI-based API with clean routing and dependency injection
• SQLAlchemy ORM for relational database management
• Role-based access with admin and normal user separation
• AWS S3 storage for image upload management
• Stripe payment endpoints for checkout and purchase flows

Getting started

Prerequisites

• Python 3.11+
• A database connection via DATABASE_URL (SQLite, PostgreSQL, etc.)
• Google API key for Gemini image generation
• Stripe secret key and optional PRICE_ID
• AWS credentials and S3 bucket configuration for file uploads

Install dependencies

python -m venv .venv
.\.venv\Scripts\activate
pip install -r requirements.txt

Configure environment

Create a .env file in the repository root with the required settings:

DATABASE_URL=sqlite:///./app.db
SECRET_KEY=your-secret-key
ALGORITHM=HS256
ACCESS_TOKEN_EXPIRE_MINUTES=60
OPEN_AI=your-openai-key
GOOGLE_API_KEY=your-google-api-key
AWS_REGION=us-east-1
AWS_ACCESS_KEY_ID=your-aws-key
AWS_SECRET_ACCESS_KEY=your-aws-secret
S3_BUCKET_NAME=your-s3-bucket
SECRET_KEY_STRIPE=your-stripe-secret-key
PRICE_ID=your-stripe-price-id
WOMPI_PUBLIC_KEY=your-wompi-public-key
WOMPI_PRIVATE_KEY=your-wompi-private-key

Note: There is no .env.example file included, so create .env manually.

Run the application

uvicorn app.main:app --reload

Open the local API docs:

• Swagger UI: http://127.0.0.1:8000/docs
• ReDoc: http://127.0.0.1:8000/redoc

Docker support

A docker-compose.yml file is present for containerized startup. Configure environment values before running.

Key features and API surface

AI chatbot & agent

• POST /user/agent — authenticated agent endpoint for clinic chatbot interactions

Image generation and editing

• POST /ai-simulation/simulate/image — generate a simulation image from an uploaded photo
• POST /ai-simulation/edit_simulated_image — edit an existing generated image using prompt input
• POST /ai-simulation/in_out_img_store — store generated simulation results for a user

Authentication and admin management

• POST /auth/login — admin login
• POST /auth/token — admin OAuth2 token endpoint
• GET /admin/admin_dashboard — admin overview metrics
• GET /admin/search_leads — search leads and simulation records

Payments and checkout

• POST /stripe/create-checkout-session — create a Stripe checkout session
• GET /stripe/success — payment success callback
• GET /stripe/cancel — payment cancel callback

Project structure

• app/ — main application package
• app/api/ — FastAPI routers for auth, admin, agent, simulation, payments, and health
• app/core/ — configuration, logging, and application bootstrap
• app/db/ — SQLAlchemy engine, session, and database dependencies
• app/models/ — ORM model definitions
• app/schemas/ — request and response schemas
• app/services/ — security, authentication, and application business logic
• app/agents/ — RAG and agent orchestration code
• app/storage/ — S3 upload utilities
• app/ml/notebooks/ — AI experimentation and demo notebooks
• docs/ARCHITECTURE.md — architecture notes and refactor roadmap
• tests/ — test suites and integration tests

Where to get help

• Use docs/ARCHITECTURE.md for architecture and migration guidance.
• Inspect app/api/ for route behavior and app/services/ for core logic.
• Run the app locally and use FastAPI documentation at /docs.

Contribution

• Fixes, enhancements, and AI pipeline improvements are welcome.
• Open issues for new features or backend changes.
• Reference docs/ARCHITECTURE.md before reorganizing major components.
• Keep new API endpoints in app/api/ and business logic in app/services/.

Maintainers

• Primary maintainer: Md Raisul Islam Rimon

Notes

• app/main.py is the preferred structured entrypoint.
• new_main.py and legacy root modules are preserved for compatibility.
• Production deployments should transition from Base.metadata.create_all() to Alembic migrations.