Nchem Takol

AI Evaluation and Prompt Engineering professional with 12+ years of experience in healthcare quality, compliance, risk management, and AI evaluation. Experienced in assessing LLM outputs, engineering and testing prompts, developing ground-truth datasets, creating evaluation rubrics, and designing test scenarios to measure model accuracy, quality, safety, consistency, and compliance. Hands-on AI Evaluator, Rater, Analyst, and Prompt Engineer skills on evaluating AI responses for factuality, relevance, completeness, reasoning, instruction following, hallucination, bias, and safety.

Experienced in identifying model failure patterns, applying objective scoring methodologies, conducting comparative evaluations, and providing actionable feedback for model and prompt improvement. Combines healthcare domain expertise in clinical documentation, care planning, HIPAA, CMS, HEDIS, and STAR with AI governance knowledge including NIST AI RMF, ISO/IEC 42001, SR 11-7, ISO 27001, and EU AI Act.

AI Evaluator | Rater | Analyst | Prompt Engineer - MERCOR / GENERALIST EXPERT

(2025)

Evaluate and rate AI-generated responses across complex tasks using structured criteria, reference materials, ground truth, and task-specific evaluation methodologies.

• Assess LLM outputs for accuracy, factuality, relevance, completeness, consistency, reasoning quality, safety, and instruction adherence.
• Design, test, and refine prompts for complex, multi-step, and context-sensitive tasks.
• Develop ground-truth answers, evaluation rubrics, scoring criteria, and validation logic for repeatable model assessment.
• Identify model weaknesses including hallucinations, unsupported claims, omissions, contradictions, reasoning errors, and instruction-following failures.
• Conduct comparative, rubric-based, and scenario-driven evaluations to measure response quality.
• Perform edge-case testing involving ambiguous, conflicting, and complex requirements.
• Analyze evaluation results to identify recurring failure patterns and opportunities for prompt and model improvement.
• Provide structured evaluation rationale and human-in-the-loop judgment for nuanced AI tasks.

QIDP / Quality Assurance Lead – AI Governance & Compliance - FESCUM INC. - Maryland

(2023 - 2025)

Led quality assurance, compliance, evaluation, and governance activities in a regulated mental health and human-services environment, developing methodologies transferable to healthcare AI evaluation.

• Designed AI-aligned evaluation frameworks and compliance scoring systems, improving audit readiness and program accountability by 30%+.
• Converted HIPAA, CMS, healthcare quality, care-delivery, and documentation requirements into measurable evaluation criteria.
• Created 5+ real-world evaluation scenarios addressing care-plan accuracy, regulatory compliance, documentation quality, and care coordination.
• Built QA tools to measure accuracy, completeness, consistency, regulatory adherence, and effectiveness.
• Conducted quantitative and qualitative evaluations to identify performance gaps and compliance risks.
• Established repeatable QA processes applicable to AI output validation and human-in-the-loop evaluation.
• Led risk identification and mitigation activities to strengthen governance and reduce compliance gaps.
• Standardized evaluation documentation, evidence collection, findings, and remediation tracking.
• Delivered executive-level quality and risk reporting to support operational and policy decisions.
• Improved compliance tracking and reporting efficiency by 25–30%.

PhD Candidate - Human Services & Public Health - Walden University

M.S. - Human Services & Public Health - Walden University

Certificate - Public Health - Harvard School of Public Health