Staff Software Development Test Engineer - AI Evaluation

Staff Software Development Test Engineer - AI Evaluation

About the Role
We are looking for a highly motivated Senior SDET – AI Evaluation to join Tekion’s AI Platform  team. Evaluation is the backbone of trustworthy AI: as Tekion scales from a handful of AI agents
to 100+ across Service, Sales, F&I, and Analytics, this role builds the evaluation platform and frameworks that let every ML team measure, trust, and improve the quality of AI outputs.In this role, you will be responsible for defining and building Tekion’s AI evaluation capabilities as a shared platform service. You will work closely with ML Engineers, Data Scientists, the AI Platform team, and Product Management to design evaluation datasets, automated scoring pipelines, and quality metrics that quantify the accuracy, consistency, and safety of AIgenerated outputs across the organization. You will own the systems that answer “is this model or agent good enough to ship, and is it 
staying good in production?” — from offline benchmarks and LLM-as-judge pipelines to online evaluation and continuous quality monitoring. You will also use AI and LLMs to scale evaluation itself, building automated judges and synthetic datasets that expand coverage faster than manual review ever could.

What You’ll Do

• Develop a deep understanding of Tekion’s AI agents, ML models, and the quality dimensions that matter for each business domain.
• Design, enhance, and own Tekion’s AI evaluation infrastructure as a shared capability used across ML teams.
• Create, curate, and maintain evaluation datasets (evals) and golden/ground-truth sets across use cases and domains.
• Define quality metrics for AI outputs — accuracy, relevance, faithfulness/groundedness, consistency, safety, and task success.
• Build automated scoring pipelines, including LLM-as-judge, rubric-based, and referencebased evaluation methods.
• Validate user intents and measure response accuracy and consistency for AI-powered capabilities such as the Analytics Agent.
• Identify hallucinations, unsafe or biased outputs, and edge cases; design targeted eval suites to catch them.
• Build both offline evaluation (pre-release benchmarking) and online evaluation (production quality monitoring, A/B, drift detection).
• Establish evaluation gates in CI/CD so model, prompt, or data changes are quality-checkedbefore release.
• Develop dashboards and reporting that make AI quality visible and actionable for ML and product teams.
• Use AI/LLMs to scale evaluation — automated judges, synthetic data generation, and eval \tooling
• Champion evaluation and responsible-AI quality best practices across the organization.

What We’re Looking For

• 5–8 years in SDET, quality engineering, ML engineering, or data science, with hands-on experience building evaluation or measurement systems — or a strong SDET background with deep LLM/ML fluency.
• Strong programming skills in Python, with the ability to build robust, reusable evaluation pipelines and tooling.
• Deep understanding of ML/LLM evaluation, benchmark design, and the pitfalls of evaluating non-deterministic systems.
• Hands-on experience with LLM-as-judge, rubric-based scoring, or human-in-the-loop evaluation.
• Solid grasp of LLM/agent concepts — prompting, RAG, embeddings, tool use — and generative failure modes (hallucination, drift, prompt sensitivity, bias).
• Experience designing and curating datasets, including labeling/annotation strategy and data quality.
• Strong statistical intuition for interpreting eval results and significance.
• Excellent communication skills to translate quality signals into decisions for ML and product teams.

Nice to Have
• Experience with eval frameworks/tools such as Ragas, DeepEval, LangSmith, TruLens, Promptfoo, HELM, or provider eval suites.
• Experience building online evaluation, guardrails, or production model monitoring.
• Familiarity with responsible AI / safety evaluation and red-teaming.
• Experience with experiment tracking (MLflow, Weights & Biases) and A/B testing.
• Prior work standing up evaluation as a platform capability for multiple teams

Experience Level

Senior Level