Decision owner
Finance analyst
The named role remains accountable for the consequential decision.
Nonprofit & Social Impact
Designed for Program director
Let the program director produce an impact report that ties indicator results to the program's logic model, reconciles outcome data with financial spend, and pairs quantitative results with consented participant stories, without overstating attribution.
Contract NON-03 · primary aggregate ProgramImpactPeriod · mode loading
Decision brief
This brief starts with the decision, evidence, and accountable owner. The guided workflow then shows how the recommendation was produced.
The evidence and control points that shape this recommendation.
Decision owner
Finance analyst
The named role remains accountable for the consequential decision.
Governed evidence
6 record types
Structured records and cited text ground the recommendation.
Workflow
5 controlled stages
Each transition records its input, output, status, and audit context.
Human controls
2 approval gates
Approval gates prevent the copilot from making consequential decisions.
A workforce program closes its fiscal year reporting 412 participants placed against a 450 target, with a 71 percent six-month retention rate carrying a medium data quality rating because 38 participants were unreachable at follow-up.
Let the program director produce an impact report that ties indicator results to the program's logic model, reconciles outcome data with financial spend, and pairs quantitative results with consented participant stories, without overstating attribution.
approvedImpactReport gives Program director the evidence needed to begin the final governed business outcome, with the decision trail retained for review.
Step 1 of 5
| Field | Value |
|---|
| Field | Value |
|---|
The output appears here immediately. More technical evidence remains available below.
The demonstration keeps deterministic controls, AI assistance, and approval separate.
Source systems enforce identity, workflow order, policy filters, and record updates.
Retrieval and generation assemble relevant evidence into a cited recommendation or draft.
Finance analyst verifies material claims and approves the consequential action.
Run the stages in order before using the Copilot. Every stage explains what happens, why it matters, the evidence it uses, and the complete API request and response.
Automated stage
What happens: Data connectors with data steward performs indicator data consolidation and quality rating using IndicatorResult and produces ratedIndicatorSet. The transition is deterministic and does not require autonomous model approval.
Why it matters: ratedIndicatorSet gives Finance analyst the evidence needed to begin Financial reconciliation to general ledger, with the decision trail retained for review.
Loading relevant sample records.
Ready to run.
POST /api/domain/workflow/indicator-data-consolidation-and-quality-rating
Run this API to see its formatted outcome.
| Field | Value |
|---|
| Field | Value |
|---|
Human approval required
What happens: Finance analyst performs financial reconciliation to general ledger using FinancialAllocation and produces reconciledSpend. A named human approver must review the evidence before the transition completes.
Why it matters: reconciledSpend gives Impact reporting copilot the evidence needed to begin Story selection under consent scope, with the decision trail retained for review.
Loading relevant sample records.
Ready to run.
POST /api/domain/workflow/financial-reconciliation-to-general-ledger
Run this API to see its formatted outcome.
| Field | Value |
|---|
| Field | Value |
|---|
Automated stage
What happens: Impact reporting copilot performs story selection under consent scope using ParticipantStory and produces eligibleStorySet. The transition is deterministic and does not require autonomous model approval.
Why it matters: eligibleStorySet gives Impact reporting copilot the evidence needed to begin Report synthesis with attribution caveats, with the decision trail retained for review.
Loading relevant sample records.
Ready to run.
POST /api/domain/workflow/story-selection-under-consent-scope
Run this API to see its formatted outcome.
| Field | Value |
|---|
| Field | Value |
|---|
Automated stage
What happens: Impact reporting copilot performs report synthesis with attribution caveats using EvaluationChunk and produces ImpactReportDraft. The transition is deterministic and does not require autonomous model approval.
Why it matters: ImpactReportDraft gives Program director the evidence needed to begin Program director and funder-variant approval, with the decision trail retained for review.
Loading relevant sample records.
Ready to run.
POST /api/domain/workflow/report-synthesis-with-attribution-caveats
Run this API to see its formatted outcome.
| Field | Value |
|---|
| Field | Value |
|---|
Human approval required
What happens: Program director performs program director and funder-variant approval using ImpactReportDraft and produces approvedImpactReport. A named human approver must review the evidence before the transition completes.
Why it matters: approvedImpactReport gives Program director the evidence needed to begin the final governed business outcome, with the decision trail retained for review.
Loading relevant sample records.
Ready to run.
POST /api/domain/workflow/program-director-and-funder-variant-approval
Run this API to see its formatted outcome.
| Field | Value |
|---|
| Field | Value |
|---|
Run a stage to see its explanation, evidence, full request, and full response.
Ask a question and receive an answer grounded in retrieved ProgramImpactPeriod evidence.
Select a starter question or enter your own.
Loaded live from /api/config, /api/domain/model, and
/api/domain/sample.
programImpactPeriods/tenantId + /programIdA workforce program closes its fiscal year reporting 412 participants placed against a 450 target, with a 71 percent six-month retention rate carrying a medium data quality rating because 38 participants were unreachable at follow-up. Spend reconciles to 1.94 million dollars across three funding sources, two restricted, yielding a 4,709 dollar cost per placement. The copilot selects four consented stories, excluding two whose consent covers internal use only, and retrieves the external evaluation's methodology note stating the absence of a comparison group. The approved funder variant reports the retention figure with its denominator and follow-up gap footnoted and states contribution rather than attribution.
| Container | Role | Hierarchical partition key | Document types |
|---|---|---|---|
programImpactPeriods |
Primary | /tenantId + /programId |
Program, IndicatorResult, FinancialAllocation, ParticipantStory, ImpactReportDraft |
evaluationLiteratureArchive |
Reference | /tenantId + /evalDocId |
EvaluationChunk |
| Type | Container | Partition key | Identifier pattern | Embedding fields |
|---|---|---|---|---|
Program |
programImpactPeriods |
/tenantId + /programId |
program::{programId} |
None |
IndicatorResult |
programImpactPeriods |
/tenantId + /programId |
program::{programId}::indicator::{indicatorCode}::{periodEndUtc} |
None |
FinancialAllocation |
programImpactPeriods |
/tenantId + /programId |
program::{programId}::financial::{periodEndUtc}::{fundingSourceCode} |
None |
ParticipantStory |
programImpactPeriods |
/tenantId + /programId |
program::{programId}::story::{storyId} |
storyVector |
EvaluationChunk |
evaluationLiteratureArchive |
/tenantId + /evalDocId |
evaluation::{evalDocId}::{chunkOrdinal} |
chunkVector |
ImpactReportDraft |
programImpactPeriods |
/tenantId + /programId |
program::{programId}::report::{periodEndUtc}::{reportId} |
None |
Compare actuals to targets across the logic model for the reporting period.
programImpactPeriods
single-partition
SELECT * FROM c WHERE c.tenantId = @tenantId AND c.programId = @programId AND c.type = 'IndicatorResult' AND c.periodEndUtc = @periodEndUtc AND c.dataQualityRating >= @minimumQuality ORDER BY c.logicModelLevel ASC
Report spend by funding source with restriction class for funder-specific variants.
programImpactPeriods
single-partition
SELECT * FROM c WHERE c.tenantId = @tenantId AND c.programId = @programId AND c.type = 'FinancialAllocation' AND c.periodEndUtc = @periodEndUtc AND c.reconciledWithGl = true ORDER BY c.spentUsd DESC
Select only stories whose consent scope permits the intended publication channel.
programImpactPeriods
single-partition
SELECT * FROM c WHERE c.tenantId = @tenantId AND c.programId = @programId AND c.type = 'ParticipantStory' AND ARRAY_CONTAINS(c.approvedUsageScope, @intendedChannel) AND c.reviewState = 'approved' ORDER BY RANK RRF(VectorDistance(c.storyVector, @themeVector), FullTextScore(c.storyText, @queryTerms))
Retrieve evaluation methodology text that bounds what the program may claim to have caused. Application-side partition-key fan-out across the evaluationLiteratureArchive reference container, bounded by the tenant prefix, the comparison-group flag and a publication lookback window, returning only the top-ranked evaluation chunks.
evaluationLiteratureArchive
partition-key-fan-out
SELECT * FROM c WHERE c.tenantId = @tenantId AND c.type = 'EvaluationChunk' AND c.comparisonGroupPresent = @hasComparisonGroup AND c.publishedAtUtc >= @evidenceLookbackUtc ORDER BY RANK RRF(VectorDistance(c.chunkVector, @queryVector), FullTextScore(c.chunkText, @queryTerms))