Umay Ana / Institutional / Pilot Package
90-DAY REFERENCE PILOT · B2B VALIDATION

Do not ask an institution to buy the vision. Give it 90 days to test the evidence.

The Umay Ana Institutional Pilot Package is a configurable reference program for banks, insurers, reinsurers, MRV providers and public institutions. The default scenario uses 10,000 invited farmers over 90 days to test adoption, field-signal generation, data quality, integration and institutional usefulness before a production-scale commitment.

Reference scenario: 10,000 invited participants · 90 days · configurable scope · commercial terms defined per partner

What the institution is testing

Whether farmer-facing utility can generate structured agricultural context that is complete enough, permissioned enough and operationally useful enough to justify production integration.

What the institution is not asked to assume

The pilot does not require the bank, insurer or MRV partner to accept an unvalidated credit model, automatic claim decision, carbon calculation or predictive risk score.

Reference package, not a promise of performance.
The 10,000-farmer / 90-day design and numerical targets below are illustrative negotiation anchors. They are not current Umay Ana performance claims and should be calibrated to the institution, country, crop mix, onboarding channel and selected use case before a pilot contract is signed.

Reference pilot design

10,000invited participants in the reference cohort
90days from activation through evaluation
5workstreams from onboarding to validation
1production-scale go / revise / no-go decision

Five phases

Pre-pilotPhase 0 · Design

Scope, cohort, consent, use case, integration and success criteria.

Days 1–15Phase 1 · Connect

Partner mapping, onboarding flows, dashboard/API sandbox and support process.

Days 16–30Phase 2 · Activate

Farmer activation, usage education and first data-quality review.

Days 31–75Phase 3 · Observe

Field-signal generation, portfolio views and institutional workflow testing.

Days 76–90Phase 4 · Evaluate

KPI review, outcome-validation analysis and scale/no-scale decision.

Workstreams

WS1 · Onboarding & Activation

Map the cohort, establish purpose/consent, activate farmers and measure funnel friction.

WS2 · Field Signal Generation

Generate crop, location, season, weather, plant-health, weekly-plan and sustainability context through useful farmer interactions.

WS3 · Data Quality & Governance

Measure completeness, provenance, coverage, permission status, evidence maturity and quality exceptions.

WS4 · Institutional Integration

Deliver agreed outputs through a pilot dashboard, secure export or the proposed institutional API contract.

WS5 · Business & Model Validation

Test whether institutional teams find the data operationally useful and identify which signals deserve outcome validation.

What success means

ActivationCan the agreed farmer cohort be mapped, consented and activated through the institution's channels?
Signal generationDo active participants generate enough crop, weather, plant-health or sustainability context to support the use case?
Data qualityAre fields complete, source-labelled, permissioned and consistently structured?
Institutional usefulnessDo analysts, underwriters or program teams repeatedly use the output in an agreed workflow?
Technical integrationCan dashboard, export or API delivery operate with traceability and agreed availability?
Outcome-validation readinessIs there enough partner outcome data to test whether selected signals justify deeper predictive validation?

Illustrative success gates

These targets make the pilot concrete enough for procurement and steering committees, while remaining negotiable. A final SOW should replace them with partner-specific thresholds.

CategoryMetricIllustrative targetContract note
OnboardingCohort mapping coverage≥95% of agreed pilot list successfully mapped or exception-codedIllustrative target — calibrate with partner before contract.
OnboardingActivated participants by Day 30≥40% of invited participantsIllustrative target — calibrate with partner before contract.
OnboardingPurpose/consent status completeness100% of externally shared recordsIllustrative target — calibrate with partner before contract.
Engagement And Signal GenerationActive participant signal coverage≥60% have at least one usable agricultural signal during pilotIllustrative target — calibrate with partner before contract.
Engagement And Signal GenerationMeaningful interactionsTarget set by use case; e.g. ≥3 structured interactions per active participant/monthIllustrative target — calibrate with partner before contract.
Engagement And Signal GenerationMulti-period coverage≥50% of active participants have signals in 2+ pilot periodsIllustrative target — calibrate with partner before contract.
Data QualityCore-field completeness≥75% on agreed institutional fieldsIllustrative target — calibrate with partner before contract.
Data QualityProvenance coverage100% of exposed fields identify source classIllustrative target — calibrate with partner before contract.
Data QualityEvidence-status coverage100% for sustainability/photo evidence recordsIllustrative target — calibrate with partner before contract.
Data QualityData coverage reported100% of dashboard/API summaries include coverage metadataIllustrative target — calibrate with partner before contract.
Institutional UsefulnessAnalyst/underwriter usefulness≥70% of pilot reviewers rate agreed views as useful or very usefulIllustrative target — calibrate with partner before contract.
Institutional UsefulnessWorkflow adoptionAt least one defined institutional workflow uses pilot output repeatedlyIllustrative target — calibrate with partner before contract.
Institutional UsefulnessDecision-boundary complianceNo automated approve/reject/pay/certify action driven solely by Umay Ana outputIllustrative target — calibrate with partner before contract.
TechnicalPilot API/export availabilityTarget ≥99.5% during agreed monitored windows, excluding planned maintenanceIllustrative target — calibrate with partner before contract.
TechnicalTraceability100% of institutional responses/export batches traceable to request/batch IDIllustrative target — calibrate with partner before contract.
TechnicalCritical security incidents0Illustrative target — calibrate with partner before contract.

Business success is not the same as predictive-model validation

A pilot can succeed operationally even before enough outcome data exists to prove a predictive relationship. For example, a bank may confirm that crop/geography/season context is valuable to relationship managers and that 65% of active farmers generate usable field signals. That does not yet prove a reduction in default.

Outcome validation is a separate workstream. Where the institution lawfully provides historical or pilot-period outcomes — such as arrears, restructuring, claims, inspections or external MRV validation — selected Umay Ana signals can be tested for association, lead time, precision and incremental information value. Only after that should the parties consider predictive or scoring claims.

First prove usefulness. Then prove predictive value.

This sequence protects both parties from turning an attractive data story into an unvalidated risk model.

Banking pilot

Inputs from bank

  • Pseudonymous pilot cohort reference
  • Agreed use case: origination or monitoring
  • Crop/geography segmentation where available
  • Existing portfolio categories
  • Optional outcome data for later validation

Outputs from Umay Ana

  • Activation and data-coverage dashboard
  • Farmer agricultural context
  • Field-signal history
  • Crop / geography / season portfolio views
  • Data-quality and provenance indicators
  • Candidate signals for later outcome validation

Insurance / reinsurance pilot

Inputs from insurer

  • Pseudonymous policy/cohort references
  • Covered crop / geography definitions
  • Selected event or monitoring use case
  • Optional claim / inspection outcomes

Outputs from Umay Ana

  • Field-risk and weather-context views
  • Plant-health observation history
  • Coverage confidence by region
  • Illustrative claim-triage context
  • Crop / regional accumulation signals
  • Validation candidates against claim outcomes

MRV / green-finance pilot

Inputs from partner

  • Selected sustainability program or methodology
  • Required evidence fields
  • Project / participant references
  • External measurement or verifier requirements

Outputs from Umay Ana

  • Carbon Diary evidence history
  • Practice taxonomy mapping
  • Evidence completeness
  • Missing-field report
  • MRV-readiness status
  • Partner-validation integration backlog

Reference integration shape

Institution cohort
Sponsored farmer access
Field interactions
Dashboard / API
Pilot KPI review
{
  "pilotRef": "pilot_bank_agri_001",
  "durationDays": 90,
  "invitedParticipants": 10000,
  "useCase": "agricultural_credit_monitoring",
  "delivery": {
    "farmerAccess": "institution_sponsored",
    "institutionalInterface": [
      "dashboard",
      "api_reference"
    ],
    "languages": [
      "partner_selected"
    ]
  },
  "successFramework": {
    "activation": "agreed target",
    "dataCoverage": "agreed target",
    "quality": "agreed target",
    "institutionalUsefulness": "reviewer score + repeated workflow usage",
    "outcomeValidation": "only where lawful partner outcome data exists"
  },
  "decisionBoundary": "Pilot measures context and operational usefulness; no autonomous credit, claim or certification decision."
}

What the institution receives at Day 90

Pilot KPI ReportActivation, engagement, data coverage, quality and workflow-use metrics.
Data Quality ReportCompleteness, provenance, evidence status, geography coverage and exceptions.
Institutional FindingsWhich views or signals were actually useful to analysts, underwriters or program teams.
Signal CatalogueWhich signal classes appeared in the cohort and at what coverage.
Integration ReviewDashboard/API/export performance and production backlog.
Outcome Validation PlanWhere partner outcome data exists, which signals merit controlled validation.
Governance ExceptionsConsent, mapping, data-quality and security issues requiring remediation.
Scale RecommendationGo / revise / no-go plus proposed production architecture and next cohort.

Commercial structure

No fictional pilot price is published in this reference package.

The commercial proposal should be built from participant count, analysis allowance, geography/language scope, onboarding work, dashboard requirements, API integration, support, retention/security requirements and any third-party MRV or verification services.

Production-scale commercial models can include institution-sponsored farmer access, per-active-farmer platform pricing, a portfolio/API subscription, enterprise licensing or a custom data/MRV integration. The final model should reflect who receives value and who carries the variable AI and support cost.

What must be agreed before Day 1

pilot cohortuse casepurpose/legal basis participant communicationlocation precisiondata fields API/export modesuccess thresholdssupport model security contactsoutcome-data availabilityDay-90 scale decision
DOWNLOADABLE PILOT ASSETS

Everything needed for a first institutional scoping meeting.

RELATED INSTITUTIONAL ASSETS

The sales narrative now has a technical execution path.

Start with 90 days. End with evidence for a scale decision.

The goal of the pilot is not to prove every long-term promise. It is to show whether Umay Ana can create enough farmer value, structured field context and institutional utility to justify the next stage.