Must-Have:

- Experience in enterprise architecture with 3+ years on Gen-AI/AI Architecture, AI governance, Responsible AI, or model-risk programs

- Experience in GenAI, Agentic AI, LLM , RAG, Azure cloud

- AI governance frameworks RBAC design, policy guardrails, audit trail configuration, and model risk concepts

- Hands-on with at least three of: Azure AI Content Safety, Azure Purview, Microsoft Defender for Cloud, Entra ID (RBAC/Conditional Access), Azure Monitor + Log Analytics, Application Insights/OpenTelemetry

- LLM observability tooling exposure: Azure AI Foundry Evaluations, or equivalent

- Reliability patterns for multi-step agent workflows: timeouts, retries, circuit breakers, idempotency keys, fallback routes, dead-letter queues

- HITL workflow design - approval workflows, escalation logic, and decision checkpoint implementation within agentic pipelines- using Microsoft Teams approvals, Adaptive Cards

- Governance documentation: runbooks, operating guides, and playbooks

Key Responsibilities:

- Design GenAI, AgenticAI architecture, workflows

- Author the governance + HITL operating model: monitoring, logging, bias/drift detection, escalation matrices, audit evidence, and runbooks

- Define RBAC (Entra ID groups, Snowflake roles), policy guardrails, and ethical safeguards across agentic workflows

- Design reliability patterns for multi-step agent workflows (timeouts, retries, and fallbacks) and reflect these in architecture artifacts and runbook guidance

- Drive deliverable acceptance through walkthroughs, dashboards, and traceable evidence within the 10-business-day Acceptance Review Period