Hr Hub
A centralized hub for AI-enabled HR workflows.
HR Hub
AI-enabled people management system. HR Hub lets an HR team manage employees, IT tasks, and support tickets through a web dashboard backed by an LLM agent that classifies requests and triggers downstream actions.
What it does
- Employee management — onboard new hires, view and edit employee records, track attrition risk scores
- IT task management — create and monitor IT tasks linked to employees
- Ticketing — submit support tickets processed by an LLM agent that classifies topics, assesses confidence, and logs actions taken
- HR chat assistant — conversational interface to the HR agent for ad-hoc queries
Project layout
| Directory | What it is |
|---|---|
backend/ |
FastAPI server — API routes, business logic, PydanticAI agent, SQLAlchemy ORM |
frontend/ |
SvelteKit dashboard — employee, task, and ticket views + chat widget |
notebooks/ |
Data work: attrition model training, DB seed notebook |
mock-cloud/ |
Local stand-in for cloud infrastructure: db/ (SQLite file), storage/ (static data files) |
Quick start
Prerequisites: uv, Yarn, Jupyter (for seeding), Docker (optional)
Development
make reset-db # delete DB → migrate → seed
make dev # backend (:8000) + frontend (:5173) concurrently
Production (local)
make prod # builds frontend, then starts both prod servers
Docker
# Copy backend/.env.example to backend/.env and fill in secrets (GROQ_API_KEY etc.)
make docker-up-build # build images and start containers
# backend → :8000 | frontend → :3000
See each subdirectory's README.md for detailed setup instructions.
Makefile targets
| Target | What it does |
|---|---|
make reset-db |
Delete DB → run revision → run migrations → seed |
make dev |
Dev servers for backend (:8000) and frontend (:5173) |
make full-reset |
reset-db then dev |
make prod |
Build frontend, start both in production mode |
make backend-prod |
FastAPI production server only |
make frontend-prod |
Build + start SvelteKit Node server (:3000) |
make docker-up-build |
Build Docker images and start all containers |
make docker-down |
Stop and remove containers |
make docker-reset |
Full wipe (volumes included) and rebuild |
