Model Evaluation Engineer

About the role

Zof AI is looking for a Model Evaluation Engineer to design and run the tests that prove AI systems actually work in real‑world settings. The role sits at the core of the product, bridging QA discipline with deep domain judgment to turn "it seems to work" into solid evidence.

Key responsibilities

• Design and build evaluation suites for AI products and agent systems.
• Develop verification harnesses that confirm the AI built the right thing.
• Define quality gates that decide what ships and what does not.
• Translate domain expertise and customer requirements into testable checks.
• Identify and hunt failure modes such as regressions, hallucinations, and silent errors.
• Make evaluation results clear and actionable for engineers, product teams, and customers.
• Integrate evaluations into CI pipelines and the development loop.
• Raise the company‑wide standard for what “working” means.

Required profile

• Experience testing, evaluating, or QA‑ing complex software systems.
• Understanding of how large language models (LLM) and agent systems can fail.
• Strong analytical rigor, skepticism, and attention to detail.
• Ability to write code for harnesses and automation.
• Clear written and verbal communication skills.
• Comfort operating in a fast‑moving environment with high ownership.

Required skills

• Software testing
• Automation
• Coding (programming) for test harnesses
• LLM evaluation
• AI verification

What we offer

• Competitive monthly salary of GHS 6,500‑13,000 plus equity.
• Full living package: housing support, private chef/meals, laundry and cleaning.
• MacBook Pro and premium AI development tools (Cursor Ultra, Claude Code Ultra, OpenAI Codex Max, etc.).
• Focused builder environment with recreational facilities (basketball court, table tennis, foosball).
• Massage, fitness, and wellness support.