Client & Professional Work

Livestock Disease Traceability Platform

A geospatial disease-traceability platform for a national livestock industry organization, integrating with federal and state animal-health systems and serving multiple distinct audiences through versioned APIs.

Highlights

  • Built a GIS-heavy Django/DRF backend with encrypted, searchable PII and full change-history auditing
  • Designed a versioned API serving three distinct audiences (partners v1/v2 and an internal UI API)
  • Delivered a role- and org-gated React frontend with behavior-driven acceptance testing throughout

Skills

Python / Django at scaleGeospatial (PostGIS / GIS)API versioning & multi-audience designData security & auditingBDD / acceptance testing

Confidentiality: Delivered under a consulting engagement. The organization's identity, the product name, and species-specific details have been generalized; named third-party government systems are described generically. Only my own work, the architecture, and the technologies are described.

Overview#

A disease-traceability platform supporting a national livestock industry organization — a geospatial system that tracks premises, animal movement, incidents, and lab results, and integrates with federal and state animal-health systems. The backend is a Django + Django REST Framework application in Python on PostgreSQL/PostGIS; the frontend is a React + Vite app whose entire experience is driven by the user's organization type and role.

The Problem#

Tracing disease across a national livestock population requires reconciling data from many organizations — producers, state veterinarians, oversight agencies — each with different permissions and views, and exchanging that data with external government systems. The platform had to model that complexity while keeping sensitive location and producer data secure and fully auditable.

My Role#

Senior Software Engineer and Tech Lead on the engagement.

Architecture & Approach#

  • Layered Django backend. The codebase is organized as vertical slices repeating across a domain app (models, business logic, serializers, validators, services), an HTTP app (DRF views, routing, auth, throttling), and a BDD app (Gherkin features with step definitions). Each domain concept — premises, incidents, movement, lab — is its own subpackage.
  • Multi-audience API. A single platform serves three distinct surfaces: a versioned, partner-facing API (v1 and v2) and a larger internal API consumed by the browser UI, using namespace versioning with auto-generated docs.
  • Role-gated frontend. A React + Vite app where routing, navigation, and visibility are composed from per-audience route modules and gated by organization type and role. Server state runs through React Query; UI/client state through Redux Toolkit — with a deliberate rule that server responses never leak into Redux.

Technical Highlights#

  • Geospatial throughout. PostGIS-backed GIS features are core to the system, with geocoding integrations and map-driven views.
  • Security and auditing. PII is stored with searchable field-level encryption; a change-history layer tracks every model mutation against the acting user; and the API stacks multiple authentication schemes (JWT, OAuth, org-scoped tokens, API keys) with granular, organization- and role-based permissions and scoped rate-limiting.
  • Behavior-driven quality. Acceptance tests run as Gherkin scenarios against an in-process integration server for partner integrations; the frontend enforces an 85% coverage gate and fails its test runs on unexpected React act() warnings or unhandled mocked requests.

Skills Demonstrated#

Production Python and Django at scale, geospatial engineering with PostGIS, multi-audience API and versioning design, field-level data security and auditing, and behavior-driven testing — a strong counterweight to the assumption that my depth is only in the TypeScript/C# world.