How admirall9/convergence works
Project type and purpose
• This is a full-stack web application combining:
• A React/TypeScript/Vite frontend (“client” folder) for interactive UIs: dashboards, budget charts, legal Q&A, citizen reviews.
• A FastAPI backend (“src” folder) providing REST/Swagger-documented APIs, data ingestion, AI services, and ETL pipelines.
• A Microsoft SQL Server database accessed via ODBC and SQLAlchemy/Alembic migrations.
• Key purpose: Serve budget transparency and legal/citizen data, support AI-driven Q&A, and periodic data ETL.
File structure
• Top-level:
• client/ → React app (components, features, UI pages, store, i18n, Tailwind).
• src/ → FastAPI app (core settings, routers, modules, services, DB models & session).
• Data_Sourcing/ → raw PDF/csv inputs.
• alembic/ → DB migration scripts.
• scripts/ → dev helpers (create-venv, migrate, run, seed_data, test_api).
• Backend structure reflects a modular monolith pattern:
• src/modules – budget, budget_etl sub-apps each with router, service, schemas.
• src/routers – API v1, health, AI, citizen, gov, legal.
• src/services – AI, ETL, legal, vector services.
• src/db/models & schemas – data definitions.
• Frontend uses feature slices and component hierarchy to mirror backend domains (budget, auth, legal, reviews, hierarchy).
README for additional insights
• FastAPI + SQL Server focus, environment variables, health checks, Alembic migrations.
• One-command dev script spins up backend and frontend.
• Testing with pytest, seed_data for initial DB content.
Detail
• Break backend into sub-components: core, routers, modules, services, models.
• Frontend high-level layers: UI pages, feature slices, store.
• Lifecycle flows: dev scripts → local env; migrations → Alembic → DB; seed_data → initial data.