← Back

Inventory Pool - A multi-tenant inventory and order-fulfillment platform for third-party logistics (3PL) operators

inventorypool screenshot

InventoryPool is a multi-tenant inventory and order-fulfillment platform for third-party logistics (3PL) operators. It's the internal system a 3PL's staff and its merchant customers would actually log into: merchants manage their own SKUs and orders, warehouse staff run the pick queue, and the platform operator bills each merchant monthly based on real usage.

Backend: Python, FastAPI, SQLAlchemy 2.0 (async, via asyncpg), Alembic migrations, Postgres 16 with native RLS. Auth via JWT (python-jose) plus Google OAuth2 (authlib) and bcrypt password hashing (passlib) offloaded to a thread pool. Billing via the Stripe SDK (test mode). Structured JSON logging (structlog) with per-request correlation IDs, metrics via prometheus-fastapi-instrumentator. Background scheduling via APScheduler (reservation expiry + monthly invoicing).Frontend: React 19 + TypeScript, Vite, React Router, TanStack Query for server state, Axios for API calls, Tailwind CSS with a small custom design-token system (colors/type/spacing as CSS variables).Infrastructure & observability: Docker Compose for local dev (backend, worker, frontend, Postgres, Prometheus, Grafana as separate services); Kubernetes manifests (namespace, StatefulSet for Postgres, Deployments for backend/worker/frontend, ConfigMaps/Secrets templates) for a kind cluster deploy. Grafana dashboards are provisioned as committed JSON tracking request latency percentiles, error rate, and a custom stock_reservation_failures_total metric.Load testing: Locust, simulating merchant order submission and warehouse pick/ship concurrently against a fixed stock pool.Tests: pytest + pytest-asyncio, including Postgres integration tests for concurrency.
Visit Live Project ↗