Build and operate production AI agents and tools on Amazon Bedrock and AgentCore. Own agent quality, retrieval, multi-step workflows, MCP endpoints, evaluation harnesses, regression gates, observability, latency, cost, and runtime guardrails. Measure groundedness, adoption, task success, and customer impact while partnering with Product and engineering teams to improve data and infrastructure. Requires strong Python, AWS, agent engineering, retrieval, evaluation, SQL, product judgment, and cross-functional communication skills.

Own the end-to-end delivery of applied machine learning products, from framing customer problems and building data pipelines to deploying, monitoring, and measuring model impact. Develop models for operational data using Python, SQL, dbt, Snowflake, and AWS; define evaluation and drift monitoring; conduct A/B tests or causal analyses; and partner with product and engineering teams to embed ML into customer workflows.