Overview
Lanexis Site Monitoring gives supervisors on construction sites a fast way to record what happened that day, and gives staff a portal to review it. The entry flow is deliberately short: select, tick, enter a quantity, take a photo, submit. I delivered it end to end as a freelance software engineer for Lanexis, and it is live for 200+ site workers.
Field workflows
- Daily progress reports and attendance for site workers.
- Material receipts, expenses, quality issues and disruptions.
- Inventory, cash balance and billing computed from the source records, never stored.
Designed for trust and bad networks
- Four roles with per-project tenant isolation: cross-project access returns 404.
- Same-day edit lock and a full audit log. There are no hard deletes, and admin corrections are preserved.
- Idempotency keys make mobile retries safe, so a flaky connection never creates duplicate entries.
- Presigned S3 URLs send photos straight to storage instead of through the API server.
Architecture
The same single-API shape as the distributor portal: thin clients on top of one Django REST backend, with all writes in a service layer and all reads through selectors scoped to the user's projects.
Portal / Supervisor app→Nginx on EC2→Gunicorn · Django 5 · DRF→PostgreSQL (RDS) + S3
Quality and documentation
- 169 pytest tests covering permissions, services and workflows.
- 6 architecture decision records and OpenAPI documentation, so the next engineer understands why, not just what.
Tech stack
The source is a private client repository. I am happy to walk through the architecture and code in an interview.