AI QE/SDET Engineer

About the Engagement This role sits within a client's spec-driven SDLC platform — an agent-first system where the specification is the single source of truth and downstream artifacts are generated from it, not described by it. The platform spans Requirements authoring, Test Data Management, CodeGen, and the QE test harness on a common foundation, built on three pillars: agent-first, input-flexible, and deterministic. Deployed on ROSA with Snowflake as the data platform and Portkey as the AI gateway. Responsibilities Re-architect the harness for headless, multi-tenant execution — durable orchestration, queued and scheduled runs Wire it into the E2E SDLC spine. Integrate with TDM's two-lane architecture and the Repo Knowledge Plane Instrument telemetry, flake detection, and stability scoring that gate promotion Experienced software engineering, recent platform/backend work Demonstrated experience re-platforming an IDE-bound tool into a shared service — core to the role Strong Python and/or TypeScript; API design, event-driven orchestration Test automation depth — Gherkin/BDD, Playwright or Selenium, CI integration Agentic/LLM frameworks (LangGraph, CrewAI, Strands) with deterministic control around probabilistic components Kubernetes/OpenShift (ROSA a plus), Snowflake, AI gateway patterns LLM cost-awareness — token economics as a design constraint Nice to Have TDM or synthetic data pipeline ownership Brownfield sequencing experience Insurance domain experience