platform engineer in reinsurance

The company’s client is an enterprise reinsurance organization building an AI enablement platform for engineering teams. The platform includes model gateways, agent and tool infrastructure, guardrails, evaluation frameworks, and AI governance standards.

• Build a model-agnostic LLM gateway on Amazon Bedrock with routing, cost attribution, and audit logging;
• Design MCP-based tool and agent infrastructure, guardrails, and retrieval patterns;
• Build evaluation frameworks for systematic AI quality measurement;
• Define production hardening standards and co-author AI governance;
• Drive organization-wide adoption of AI-native tooling, including Claude Code, Amazon Kiro, and Copilot;
• Mentor client engineers to ensure the capability remains within their teams.

• Have 10+ years of software engineering experience and recent hands-on AWS production delivery;
• Have consulting, platform, or enablement experience serving other engineering teams;
• Have introduced standards and practices that teams actually adopted;
• Have proven AI-native engineering practice, using agentic coding tools as the primary workflow;
• Have deep AWS serverless and container experience with Lambda, Step Functions, DynamoDB, ECS/Fargate, and CDK;
• Have hands-on Bedrock and LLM platform experience with gateways, prompt engineering, and evaluation;
• Have strong Python skills and solid TypeScript/Node skills;
• Understand agent and tool security, including prompt injection, PII handling, and blast radius;
• Nice to have: AWS certifications, Bedrock Agents, regulated industry background, Dremio, Iceberg, Airflow, DBT.

• 2-3 Year contract.