Engineering Change Impact Copilot
Outcome
Let the configuration manager see the true blast radius of a proposed engineering change: which requirements, parts, effectivity ranges, qualification tests and publications it disturbs, so the change board decides with the affected artifacts enumerated rather than estimated.
| Portfolio attribute | Value |
|---|---|
| Primary persona | Configuration manager |
| Primary aggregate | EngineeringChangeRequest |
| Cosmos DB containers | 2 |
| Grounding data | requirements, BOM, tests, changes, manuals |
| Development estimate | USD 60-220 monthly dev estimate |
| Runtime options | Python FastAPI and C# ASP.NET Core |
Productivity hypothesis
These measures reflect the workflow, evidence, decision boundary, and industry risk for this use case.
| Impact type | Measure | Conservative internal assumption | Customer measurement |
|---|---|---|---|
| Productivity | Task cycle time | 10-20% reduction | (baseline median task minutes - assisted median task minutes) / baseline median task minutes * 100 |
| Quality or risk | First-pass acceptance | 5-10 percentage-point improvement | assisted percentage accepted without material rework - baseline percentage accepted without material rework |
Important: Conservative internal pilot assumptions only; not public benchmarks, commitments, or expected customer results. Validate against a comparable baseline or matched control cohort with mandatory human review.
Rationale: Configuration managers can trace requirements, bills of material, tests, changes, and manuals more quickly while retaining formal approval.
Pilot guardrail: Surface missing or conflicting change-control evidence rather than resolving it through inference.
Customer validation plan
- Baseline the current workflow with representative users and tasks.
- Run the accelerator on the same task mix with human review retained.
- Measure task cycle time and first-pass acceptance with the documented formulas.
- Replace internal assumptions with customer-observed results before a business case.
Why Azure Cosmos DB
This AI solution needs an operational data plane, not only a vector index:
- Hybrid vector and full-text retrieval combines semantic recall with exact identifiers, codes, names, and domain language.
- Hierarchical partition keys and metadata predicates keep tenant and domain boundaries in the normal query path.
- Operational JSON, source chunks, workflow and agent state, feedback, and audit records remain together for low-latency reads and governed writes.
- Change feed reacts to source and operational updates without polling.
- Autoscale throughput, configurable consistency, and multi-region distribution support variable demand and enterprise availability.
Domain workflow

| Stage | Actor | Input → output | Approval |
|---|---|---|---|
| 1. Change intake and classification | Configuration manager | ChangeRequest → classifiedChange | Policy controlled |
| 2. Requirement and BOM traceability expansion | Traceability resolver | classifiedChange → affectedArtifactSet | Policy controlled |
| 3. Requalification determination | Engineering change impact copilot with test engineering | QualificationTestLink → requalificationVerdictSet | Policy controlled |
| 4. Publication impact scoping | Engineering change impact copilot | PublicationImpactChunk → publicationUpdateScope | Policy controlled |
| 5. Change control board disposition | Change control board | ImpactAssessment → boardDisposition | Human required |
Technical architecture

Cosmos DB data model
| Container | Hierarchical partition key | Documents | Search paths |
|---|---|---|---|
engineeringChangeRequests | /tenantId, /changeRequestId | ChangeRequest, RequirementLink, BomImpact, QualificationTestLink, ImpactAssessment | /changeVectorEmbedding, /requirementVectorEmbedding, /changeVector, /requirementVector |
technicalPublicationLibrary | /tenantId, /ataChapter | PublicationImpactChunk | /chunkVectorEmbedding, /chunkVector |
Query contract
| Query | Container | Scope | Purpose |
|---|---|---|---|
| InvalidatedVerifications | engineeringChangeRequests | single-partition | List requirements whose verification the change invalidates and that therefore need reverification. |
| FieldedUnitExposure | engineeringChangeRequests | single-partition | Quantify how many fielded units and how much existing stock the effectivity change disturbs. |
| RequalificationCostDrivers | engineeringChangeRequests | single-partition | Rank qualification tests requiring repeat or delta qualification by cost and duration. |
| AffectedPublicationSections | technicalPublicationLibrary | cross-partition-bounded | Find publication sections whose effectivity or content the change invalidates so updates are scoped before release. |
Python and C# services route each document type to its declared container and hierarchical partition key. ETag-based writes protect mutable aggregates, while append-only records and bounded cross-partition queries preserve scale.
Azure deployment

Clone the source repository, authenticate with Azure, and run:
Set-Location accelerators/aerospace-defense/engineering-change-impact-copilot
az login
./Deploy.ps1 -Runtime python -Location eastus2
Change python to csharp to deploy the .NET implementation.
Explore the implementation
- Launch signed-in demo
- Implementation article
- Domain data model
- Workflow and API contract
- Python application
- C# application
- Domain container Bicep
- Platform Bicep
- Cost and configuration
- Source repository