VeldWatch

Rustlers Valley Farm is an Eastern Free State regenerative livestock farm which ordinally rotates cattle, chickens and pigs, through 75 grazing camps so that each camp can rest and the soil recover. This model depends on proof: certification bodies, investors and directors need evidence of rotation adherence. It is coordinated through WhatsApp and paper records, so the farm manager struggles with the retrieval process and data storing making it prone to human error and duplication. Additionally, there is no audit trail, directors cannot see conditions in real time, and anyone relaying a message can effectively authorise a herd move and animals' proneness to overgrazing increases.

This is where VeldWatch comes in, replacing informal messaging with one role-aware system for three users: workers who need quick capture without signal, a manager who needs one view of all  camps, and directors who need plain-language evidence of land improvement. A React dashboard and offline-capable React Native app sit on a three-tier architecture: an ASP.NET Core Web API (.NET 10) and a 14-table Supabase PostgreSQL/PostGIS database in Third Normal Form, with Mapbox and Telegram integrations.

These features answer the problems. On the mobile app, workers log herd entries, exits and grass observations in English or Sesotho; entries queue offline and sync when signal returns, while validation enforces rotation order, blocks entry to recovering camps and audit-logs every override. On the web dashboard, managers and directors see a live Mapbox map of all 75 camps coloured by status and herd, driven by Sentinel-2 NDVI scores refreshed every five days. Telegram alerts, role-based access and plain-language reports keep managers informed and give directors evidence to share.

VeldWatch is published and being refined, with most features built and awaiting demo testing for farm feedback. We conclude that one auditable source of truth is achievable within the farm's constraints.