AI Verification Engineer (Senior Full-Stack)

Windmill Digital, за кордоном

Що робити

Review and verify code produced by AI coding agents — assessing correctness, architecture, performance, and security to the standard of a senior engineer, not surface-level testing.
Own quality end-to-end before releases ship: deep manual, exploratory, and scenario-based testing across both AI and non-AI features, plus fix verification and regression prevention.
Evaluate AI-powered features — LLM-based flows, recommendation engines, decision logic — validating outputs for correctness, consistency, bias, hallucination risk, and edge cases, with clear qualitative and quantitative acceptance criteria.
Identify non-determinism, prompt sensitivity, and reliability risks specific to AI systems, and propose practical mitigations.
Build and maintain automated evaluation and regression pipelines for AI-enabled systems — using AI agents (Cursor, Claude Code) in your own verification workflow.

AI-Native Development: Cursor or Claude Code (or equivalent agentic tools); prompt engineering, applied to both building and verifying
Engineering depth: 8+ years across the full stack, strong enough to review and validate another engineer’s (or AI’s) code
Front-End: React or Angular, with TypeScript / JavaScript
Back-End: Java, Node.js, or Ruby
AI system behavior: solid understanding of non-determinism, prompt sensitivity, and the challenges of validating model outputs

Знайдено через trackr.help/jobs · Канал: @trackrhelp · Бот для персональних сповіщень: @trackrhelpBot