Transparency
Methodology
Every number on the platform is computed from a documented formula. Below is how the KPI Engine, Benchmark Engine and Nutrition Economics Engine work end-to-end.
1. Raw Data Layer
Production records (entries, deaths, weights, feed deliveries, milk drops, eggs) and financial records (feed costs, sale prices, fixed costs) flow into typed tables with full audit history.
2. KPI Engine
Species-specific functions (
swine-kpis, poultry-kpis, dairy-kpis) compute ADG, FCR, EPEF, Livability, Milk Yield, SCC, IOFC and 30+ other metrics directly from records. Each computation is versioned as a KPI snapshot.3. Benchmark Engine
Industry Average, Top 25%, Top 10% and Elite bands are computed from the contributing cohort (anonymized) plus peer-reviewed regional/global research. World Class is set at +5% beyond Elite for higher-is-better metrics and −5% for lower-is-better metrics.
4. Nutrition Economics Engine
- Feed Cost/kg Gain = feed cost ÷ live-weight gain
- IOFC = revenue − feed cost
- IEVI™ — ranks each ingredient on protein, energy, AA contribution and price
- Sensitivity — tornado simulation across ±20% ingredient price moves
5. AI Analysis Engine
Gemini-class model receives the structured KPI + benchmark + economics snapshot and produces an executive summary, root-cause analysis, and a 30/60/90-day action plan. Outputs are cached in
ai_analyses for replay and audit.6. Audit Engine
Mobile-first checklists with weighted scoring. Each finding is risk-ranked (low/medium/high/critical) and rolled up into a single audit score per farm.