Proposed data handoff and buyer validation

This is a pilot design, not an implemented integration or a partner agreement.

Before season one, the cooperative and prospective reviewer would agree one evidence schema: plot and season identifiers; crop dates; water observations with time, source and quality flags; nitrogen and residue inputs; yield; reviewer decisions; calculation version and assumptions. Raw observations and later corrections would remain distinguishable. Unresolved cases would be reported explicitly.

The pilot would compare the existing process with LuaZero on reviewer minutes per completed package, completeness, number of corrections, reasons for rejected packages, and field support cost. Both accepted and rejected cases must remain in the evaluation denominator. A buyer's documented renewal decision would complement these operational measures.

An agreement must define data ownership, farmer consent, permitted access, retention, cost sharing and rights to claim outcomes before enrolment. A data export is not a transfer of carbon rights or proof that two organisations can claim the same result. National programme reporting and voluntary certification require separate mappings.

The current prototype exports synthetic demonstration JSON and CSV. No external provider has accepted those exports, and no integration with Regrow or another MRV provider is claimed.