Principal QA Engineer

Country

United States

State

NA

Date Published

06-Oct-2026

Job

47645

Travel

You may occasionally be required to travel for business

Additional Locations

This role can be based remotely in United States

BMC empowers nearly 80% of the Forbes Global 100 to accelerate business value, faster than humanly possible. Our industry-leading portfolio unlocks human and machine potential to drive business growth, innovation, and sustainable success. BMC does this in a simple and optimized way by connecting people, systems, and data that power the world’s largest organizations so they can seize a competitive advantage.

The IZOT product line includes BMC’s Intelligent Z Optimization & Transformation products, which help the world’s largest companies to monitor and manage their mainframe systems. The modernization of mainframe is the beating heart of our product line, and we achieve this goal by developing products that improve the developer experience, the mainframe integration, the speed of application development, the quality of the code and the applications’ security, while reducing operational costs and risks.

We acquired several companies along the way, and we continue to grow, innovate, and perfect our solutions on an ongoing basis.

You build the systems that tell us whether our agentic AI is good enough to ship: evaluation frameworks, golden datasets, and rubrics; automated eval harnesses in CI; and drive down failure modes such as hallucination, drift, and unsafe or non-repeatable output. You are the reason customers can trust what our agents produce. At this level you independently own well-scoped work from definition through delivery.

Scope at this level:
Independently owns well-scoped eval work for a feature/team from definition through delivery. Organizational impact expected:
Improves outcomes for one team or project.

• Define and implement evaluation strategies and frameworks for Generative AI, RAG, and agentic AI applications.
• Create golden datasets, rubrics, and scoring methods (including LLM-as-a-judge) so quality is measured consistently.
• Apply statistical sampling, repeated trials, judge calibration, dataset versioning, and anti-contamination practices.
• Stand up automated evaluation harnesses in CI so model/prompt/agent changes are re-scored before ship.
• Define quality metrics and release gates for customer-ready vs prototype.
• Design testing strategies for multi-agent workflows and AI systems using tools, APIs, MCP, and enterprise apps;
  Playwright where relevant.
• Evaluate end-to-end agent behaviour (planning, tool use, response generation, task completion).
• Identify failure modes using observability/tracing; recommend reliability improvements.
• Close the loop from evaluation results into model/prompt improvements with DS and AI Engineering.
• Champion Responsible-AI and human-in-the-loop principles; contribute to governance and model validation.

• Bachelor's or master's in computer science, AI, or related field (or equivalent experience).
• Experience with Generative AI / LLM applications or AI/ML systems and quality/evaluation frameworks (depth scales with level).
• Understanding of LLM/RAG/agentic failure modes; vector DBs and embedding retrieval as level requires.
• Python; building eval sets, rubrics, automated scoring; CI/CD integration.
• LLM observability/tracing (e.g. Langfuse, OTel) and analytics (Open Search or similar) as level requires.
• Rigorous, data-driven mindset; collaboration across DS, engineering, and product.
• Evidence of owning an eval suite that influenced a release decision.
• Builds consistent rubrics/datasets — not one-off manual spot checks only.
• Works…