One platform, four capabilities
Phoeniks turns public farmland listings and open Earth-observation data into per-parcel yield estimates that carry their own audit trail. The product is organised around four capabilities — parcel search, yield prediction, evidence and diligence, and data access — sitting on three internal layers: Distribution, Intelligence and Infrastructure.
The three layers behind the product#
The platform is engineered in three layers. Distribution is what a buyer or broker touches: listings, maps, filters, diligence packs. Intelligence is the core IP: a crop-model ensemble producing per-property yield estimates and confidence. Infrastructure is the part institutions audit: ingestion, compute, provenance and evidence governance.
| Layer | What it does | Who relies on it |
|---|---|---|
| Distribution | Searchable farmland inventory, maps, filters, diligence checkout, buyer and broker workspaces | Brokers, buyers, agencies |
| Intelligence | APSIM / DSSAT / ORYZA ensemble, crop-fit signal, scoring engine, validation packs, institutional feed | Banks, reinsurers, allocators |
| Infrastructure | Satellite and weather ingestion, geocoding, batch compute, provenance ledger, evidence governance | Model risk, audit, compliance |
What makes it defensible#
- Data-neutral modelling: no dependency on a single satellite or weather provider.
- Every output is reproducible from a versioned model, a versioned input set and a recorded scenario.
- Evidence scoring levels let a reviewer see what is measured, what is inferred and what is untested.
- Compliance is engineered into the pipeline, not bolted on: 188 of 319 platform capabilities carry a high or critical regulatory rating.
Frequently asked
- Is Phoeniks a tool for farmers?
- No. Phoeniks is institutional. Its users are lenders, reinsurers, allocators, brokers and farmland buyers who need defensible yield evidence for a decision, not in-field agronomy advice.
- Does Phoeniks need on-farm sensors?
- No. The platform runs on public and licensed sources — satellite imagery, reanalysis weather, open parcel registries and official yield statistics — so a property can be assessed before any site visit.
Related pages
Parcel search
Search a normalised farmland inventory built from public portals, with map results, filters, source provenance on every listing and shortlist comparison.
Yield prediction
An APSIM, DSSAT and ORYZA ensemble producing per-property yield estimates, crop-fit signals, scenario runs and a transparent scoring engine.
Evidence and diligence
Validation packs, evidence scoring levels, a reproducibility ledger, diligence PDF export and a pre-diligence room for reviewable farmland decisions.
Data and API
Sentinel-1, HLS, AgERA5, CHIRPS, France RPG, IBGE SIDRA and CIMMYT ingestion, exposed through a listing intelligence REST API and an institutional feed.
Carbon MRV evidence
Baseline reconstruction, monitored change with uncertainty and permanence-risk inputs for agricultural carbon projects and buyers.
Insurance analytics
Yield deviation, basis-risk analysis and parametric trigger design built on documented crop models and replayable weather scenarios.
Scope 3 agriculture
Sourcing-region evidence for food and CPG companies reporting agricultural Scope 3 emissions and land-use change.
Commodity intelligence
Regional production trajectories and in-season deviation signals for traders and analysts covering agricultural commodities.
Data sources
The families of data Phoeniks ingests: satellite, weather and reanalysis, soil, land registry, official statistics and market records.
See also
Request access
Institutional access. No self-serve trial.