How pacharaponk/aphrodize works
Aphrodize combines a user-facing skin and wellness application with a modular ML workflow backend. The image-analysis path runs from capture through API submission, queued inference, the checked FFHQ-Wrinkle pipeline, and persisted results; account, product recommendations, UV, and Daily Health are distinct product capabilities. The README and source index support these branches, though several route and model implementations were not sampled, so uncertain connections are omitted.
Main entry points: router.py Versioned API router · inference_worker.py Inference worker · analysis_service.py Analysis workflow · service.py Wrinkle orchestration
Components 33
User application
- Image capture
frontend/src/app/capture/page.tsx - Web API routes
frontend/src/app/api/analysis/route.ts - Analysis results
frontend/src/app/capture/analysis-result.tsx - Recommendations UI
frontend/src/app/recommendation/recommendation-panel.tsx - UV experience
frontend/src/app/uv-map/page.tsx - Daily Health UI
frontend/src/app/clients/daily-health-dashboard.tsx
API and workflows
- FastAPI application
backend/main.py - Versioned API router
backend/api/v1/router.py - Accounts and consent
backend/api/v1/routes/auth.py - Image analysis routes
backend/api/v1/routes/analyses.py - Generic inference route
backend/api/v1/routes/inference.py - Product catalog routes
backend/api/v1/routes/products.py - UV routes
backend/api/v1/routes/uv.py - Analysis workflow
backend/services/analysis_service.py - Inference enqueue
backend/services/inference_service.py - UV services
backend/services/uv_service.py - UV map service
backend/services/uv_map_service.py
Image inference
- Inference worker
backend/workers/inference_worker.py - Wrinkle orchestration
backend/wrinkle/service.py - Prediction pipeline
ai/ffhq_wrinkle/prediction.py - Model loading
ai/ffhq_wrinkle/modeling.py - Segmentation models
ai/ffhq_wrinkle/official/inference.py - Confidence policy
ai/ffhq_wrinkle/confidence.py - Wrinkle scoring
ai/ffhq_wrinkle/scoring.py
Platform services
- Application database
backend/core/db/models.py - Private object storage
backend/libs/minio_client.py - ARQ job queues
backend/libs/redis_client.py - Annotation workflow
backend/services/annotation_service.py
Daily Health
- Daily Health routes
backend/api/v1/routes/daily_health.py - Daily Health services
backend/services/daily_health_dataset.py - Daily Health training
backend/services/daily_health_training.py
Other components
- User
- Label Studio
Connections 41
- User → Image capture: submits photo
- Image capture → Web API routes: sends analysis
- Web API routes → Image analysis routes: calls API
- Image analysis routes → Analysis workflow: creates analysis
- Analysis workflow → Application database: stores workflow
backend/services/analysis_service.py - Analysis workflow → Private object storage: reads and writes
backend/services/analysis_service.py - Analysis workflow → ARQ job queues: queues job
backend/services/analysis_service.py - ARQ job queues → Inference worker: dispatches job
backend/workers/inference_worker.py - Inference worker → Application database: loads and saves
backend/workers/inference_worker.py - Inference worker → Private object storage: loads image
backend/workers/inference_worker.py - Inference worker → Wrinkle orchestration: runs analysis
backend/workers/inference_worker.py - Wrinkle orchestration → Prediction pipeline: predicts image
backend/wrinkle/service.py - Wrinkle orchestration → Confidence policy: applies policy
backend/wrinkle/service.py - Wrinkle orchestration → Wrinkle scoring: scores output
backend/wrinkle/service.py - Prediction pipeline → Model loading: loads model
ai/ffhq_wrinkle/prediction.py - Model loading → Segmentation models: constructs network
ai/ffhq_wrinkle/modeling.py - Inference worker → Private object storage: stores artifacts
backend/workers/inference_worker.py - Inference worker → Annotation workflow: stages review
backend/workers/inference_worker.py - Annotation workflow → Label Studio: publishes tasks
backend/services/annotation_service.py - Image analysis routes → Analysis results: returns analysis
- User → Accounts and consent: creates account
- FastAPI application → Versioned API router: mounts routes
backend/main.py - Versioned API router → Accounts and consent: registers routes
backend/api/v1/router.py - Versioned API router → Image analysis routes: registers routes
backend/api/v1/router.py - Versioned API router → Generic inference route: registers routes
backend/api/v1/router.py - Versioned API router → Daily Health routes: registers routes
backend/api/v1/router.py - Versioned API router → Product catalog routes: registers routes
backend/api/v1/router.py - Versioned API router → UV routes: registers routes
backend/api/v1/router.py - Generic inference route → Inference enqueue: creates run
backend/api/v1/routes/inference.py - Inference enqueue → Application database: persists run
backend/services/inference_service.py - Inference enqueue → ARQ job queues: enqueues job
backend/services/inference_service.py - User → Daily Health UI: tracks health
- Daily Health UI → Daily Health routes: requests data
- Daily Health routes → Daily Health services: uses services
- Daily Health training → Application database: records training
- User → Recommendations UI: views products
- Recommendations UI → Product catalog routes: requests catalog
- User → UV experience: views UV
- UV experience → UV routes: requests UV data
- UV routes → UV services: uses service
- UV routes → UV map service: uses map service