Camunda Architect

1.Define the GB Camunda contract-usability model up front: what constitutes a process instance, how long-running claims are decomposed into modular orchestrated events/sub-processes, and how we avoid designing one monolithic process per claim.

2. Define the headless target architecture with GB Navigator remaining the adjuster experience and Camunda operating behind it as the durable orchestration tier, keeping process definitions portable through BPMN/DMN standards.

3. Architect the MVP event-driven abstraction/integration layer as a service-bus-style boundary between Camunda and RISX-FACS/HPE NonStop, Azure services, APIs and external vendors. Define the canonical events/contracts, correlation and idempotency model, routing, reusable adapters/job workers and error/retry patterns so Camunda never becomes directly coupled to individual systems and the same integration services can be reused by other orchestrations and applications.

4. Define the claim orchestration model: a lightweight parent claim lifecycle coordinating independently deployable workflows/sub-workflows (potentially 150+ per claim), triggered by claim events and capable of long-running, human-in-the-loop execution.

5. Establish clear separation of concerns between orchestration, business rules and implementation code: BPMN for process flow, DMN/Drools for externalized decisions, and services/workers for technical execution so rules and processes can evolve independently.

6. Define and prototype a natural-language-to-process capability grounded in the RISX-FACS knowledge layer (Atlas/Navigator, vector + graph knowledge, MCP/APIs and SME-validated rules). Establish the steps from a business user's natural-language intent through retrieval of RISX-FACS capabilities, rules, events and dependencies; generation of BPMN/DMN candidates; architecture/SME validation; automated testing; version-controlled approval; and governed deployment into Camunda. The objective is not generic AI process generation, but process creation grounded in GB's machine-understandable system and business knowledge.

7. Produce the production foundation before process build-out: reference patterns for SSO/security, correlation and idempotency, retries/error handling, versioning and migration of in-flight processes, observability/audit/SLA telemetry, CI/CD, DR/supportability, and SaaS-to-self-hosted portability.

8. Deliver an implementation-ready architecture package and prioritized MVP backlog for the first claim workflows, including standards, reusable patterns, non-functional requirements, operating model and a handoff that lets GB scale additional workflows without reinventing the platform.