Agentic AI Architect

Job title : Agentic AI Architect

Experience

12+ years, including 1+ years in LLM and agentic AI

Domain : Insurance and financial services preferred

Role summary

Design and deliver governed, agentic AI solutions that speed up software delivery for insurance and financial services clients. You will own the architecture end to end, from knowledge layer and agents to guardrails and Azure DevOps integration, and lead client engagements from discovery to handover.

Must-have skills

• Hands-on Python plus one of C#/.NET, Java or TypeScript; strong client communication
• Multi-agent orchestration, tool calling, planning and human-in-the-loop design
• RAG, hybrid search, vector stores and knowledge graphs of code and architecture
• Prompt and context engineering, LLM evaluation, token and cost management
• Azure DevOps APIs, CI/CD, branch policies and test automation
• Azure security: Entra ID, managed identities, Key Vault, private endpoints
• AI governance: guardrails, audit trails, PII handling and responsible-AI controls

Key responsibilities

• Run discovery: AI maturity and SDLC assessment, use-case selection and baseline metrics.
• Own the target architecture: agent orchestration, SDLC agents, knowledge base and knowledge graph, integration and security.
• Configure and tune agents, skills, prompts and hooks on Agentic platform; no model fine-tuning.
• Turn engineering standards and policies into guardrails as code, with human approval and full audit trails.
• Integrate agents with Azure DevOps (Boards, Repos, Pipelines, Test Plans) and measure results against the baseline.
• Lead demos, steering updates and handover; keep delivery within scope and raise changes early.
• Guide architects and engineers across onshore and offshore teams.