System record 01 / 02
Distribution Management System
A place for every device. A record of every handover.
Measured against the resume record
Full-stack build
I built the system end to end: the data model, the approval workflows and the role-scoped interface that keeps every handover accountable.
Every device is registered with a serial number and tracks its status, location and current holder. Each movement is written to a device history, so the full chain from registration to return stays auditable. Devices and users can also be bulk-imported from CSV, XLSX or XLS.
Distribution requests move through a structured approval chain of pending, approved, delivered or rejected, with approver notes attached at each decision. The same lifecycle view shows where a request stands from any role-scoped dashboard.
Defects are reported by type and severity and resolved through a history-tracked workflow. Returns are raised with reason categories and approved through the same chain and both sit beside distributions in a single approval queue.
Access is governed by a hierarchy of eight roles, from system administration through field operations. Each role gets its own dashboard and route access and a permission matrix of more than 55 entries keeps those boundaries explicit rather than implied.
Role-specific dashboards show live statistics, activity feeds, charts and system alerts. Notifications carry an unread badge and mark-as-read state and reports cover inventory, distribution, defects, returns, user activity and device utilisation.
The schema spans seven entity domains across 21 MySQL tables. An idempotent migration system keeps environments in sync without manual steps, so the data model can move forward safely as the system grows.
Prometheus scrapes the backend metrics endpoint every ten seconds while SQLAlchemy engine events record query volume, latency, failures and active connections. Grafana provisions Backend API and Database dashboards on startup.