How tri0nic/hackathon-09-26 works
This is a fire-risk monitoring and response MVP for dispatchers: users inspect forecast alerts and sensor context, review explanations and SMS status, record decisions, and create or execute preventive-maintenance requests. The React interface calls an ASP.NET API, which owns the business workflow and PostgreSQL persistence and delegates predictions to the FastAPI ML service. The ML package also contains a substantial data-preparation and model-training workflow. Unsampled routes and storage behavior are represented at subsystem level; edges are limited to relationships visible in excerpts, the source index, or README.
Main entry points: App.tsx Risk dashboard · Program.cs HTTP workflow routes · ModelDemoService.cs Demo workflow · pipeline.py Data preparation
Components 21
Dispatcher interface
- Risk dashboard
services/web/src/App.tsx - Model demo
services/web/src/ModelDemoPage.tsx - Web API client
services/web/src/api.ts - Risk domain logic
services/web/src/domain.ts - Shared UI
services/web/src/components.tsx
Workflow API
- HTTP workflow routes
services/api/FireRisk.Api/Program.cs - Business rules
services/api/FireRisk.Api/Domain/BusinessRules.cs - Demo workflow
services/api/FireRisk.Api/ModelDemoService.cs - Scenario catalog
services/api/FireRisk.Api/ModelDemoScenarioCatalog.cs - Calculation cache
services/api/FireRisk.Api/ModelDemoCalculationCache.cs - ML gateway client
services/api/FireRisk.Api/MlClient.cs
Prediction service
- FastAPI endpoints
services/ml/src/fire_risk/ml_api.py - Artifact inference
services/ml/src/fire_risk/ml/inference.py
ML data and training
- Data preparation
services/ml/src/fire_risk/data/pipeline.py - Feature engineering
services/ml/src/fire_risk/data/features.py - Data quality checks
services/ml/src/fire_risk/data/quality.py - ML data contracts
services/ml/src/fire_risk/contracts.py - Model training
services/ml/src/fire_risk/ml/training.py
Operational state
- PostgreSQL store
services/api/FireRisk.Api/PgStore.cs - PostgreSQL
Other components
- Dispatcher
Connections 22
- Dispatcher → Risk dashboard: uses
README.md - Risk dashboard → Web API client: requests data
services/web/src/App.tsx - Risk dashboard → Shared UI: renders
services/web/src/App.tsx - Risk dashboard → Risk domain logic: uses
services/web/src/App.tsx - Risk dashboard → Model demo: renders
services/web/src/App.tsx - Model demo → Web API client: requests demo
services/web/src/ModelDemoPage.tsx - Web API client → HTTP workflow routes: calls HTTP API
README.md - HTTP workflow routes → PostgreSQL store: reads and writes
services/api/FireRisk.Api/Program.cs - HTTP workflow routes → ML gateway client: invokes
services/api/FireRisk.Api/Program.cs - HTTP workflow routes → Demo workflow: routes demo
services/api/FireRisk.Api/Program.cs - HTTP workflow routes → Business rules: applies
- Demo workflow → Scenario catalog: loads scenario
services/api/FireRisk.Api/ModelDemoService.cs - Demo workflow → Calculation cache: stores calculation
services/api/FireRisk.Api/ModelDemoService.cs - Demo workflow → ML gateway client: requests prediction
services/api/FireRisk.Api/ModelDemoService.cs - ML gateway client → FastAPI endpoints: sends prediction request
README.md - FastAPI endpoints → Artifact inference: runs predictor
services/ml/src/fire_risk/ml_api.py - PostgreSQL store → PostgreSQL: persists workflow state
README.md - Data preparation → Feature engineering: builds features
services/ml/src/fire_risk/data/pipeline.py - Data preparation → Data quality checks: checks quality
services/ml/src/fire_risk/data/pipeline.py - Data preparation → ML data contracts: uses records
services/ml/src/fire_risk/data/pipeline.py - Model training → Feature engineering: trains from prepared data
- Model training → Artifact inference: produces model artifacts
services/ml/src/fire_risk/ml/inference.py