This isn't feature delivery. You'll work hands-on as a senior engineer while consulting to the client's architects and engineering teams — success is measured by how much faster, safer, and more capable they become.

What you'll do:🔹 Build a model-agnostic LLM gateway on Amazon Bedrock (routing, cost attribution, audit logging)🔹 Design MCP-based tool/agent infrastructure, guardrails & retrieval patterns🔹 Build evaluation frameworks so AI quality is measured systematically, not per-project🔹 Define production hardening standards & co-author AI governance🔹 Drive adoption of AI-native tooling (Claude Code, Amazon Kiro, Copilot) org-wide🔹 Mentor client engineers so the capability stays with them

• What you bring:✅ 10+ years software engineering, recent hands-on AWS production delivery✅ Consulting/platform/enablement experience — your customers were other engineering teams✅ A track record of standards & practices you introduced that teams actually adopted✅ Proven AI-native engineering practice — agentic coding tools as your primary workflow✅ Deep AWS serverless/container skills (Lambda, Step Functions, DynamoDB, ECS/Fargate, CDK)✅ Hands-on Bedrock/LLM platform experience — gateways, prompt engineering, evaluation✅ Strong Python, solid TypeScript/Node✅ Understanding of agent/tool security (prompt injection, PII handling, blast radius)