Thomas Ngau

Results-driven Generative AI Evaluator and Prompt Engineer with a Master's in Data Science & Artificial Intelligence from Carnegie Mellon University. Specialized in rigorously assessing LLM-generated responses for factual accuracy, truthfulness, completeness, logical depth, and hallucinations. Proven ability to design diverse, edge-case test prompts, apply structured rubrics, annotate complex reasoning traces, and deliver actionable insights that strengthen AI reliability and safety.

With 2+ years of hands-on GenAI validation experience, remote asynchronous project leadership, and a deep commitment to ethical AI, I deliver consistent, high-quality evaluations that directly elevate model performance at scale.

Freelance Generative AI Consultant & QA Analyst - Independent Contractor (Fully Remote) - Remote

(2023-06)

• Evaluated GenAI outputs across text-based tasks for startups, scoring responses on factual accuracy, insight depth, completeness, and coherence while systematically identifying and documenting hallucinations and quality gaps.
• Crafted targeted, creative prompts to probe model boundaries, exposing inconsistencies in policy logic for chat-based AI agents and increasing edge-case coverage by 25%.
• Annotated detailed reasoning paths and generated plausible human-aligned alternatives using YAML/JSON structures, enabling precise model refinement and gold-standard behavioral validation.
• Streamlined multi-project evaluation pipelines, cutting documentation and traceability time by 30% through optimized asynchronous workflows and cross-functional collaboration with developers.
• Managed concurrent client engagements across time zones, consistently meeting quality thresholds

Research Assistant - Carnegie Mellon University, Center for AI and Society - Pittsburgh, PA, USA

(2022-09 - 2023-05)

• Designed and executed structured JSON/YAML test cases for LLM-based decision-making systems, uncovering hidden assumptions and potential failure modes in autonomous AI agents.
• Co-authored "Holistic Scenario Design for AI Agents," a research paper presented at the CMU Symposium (2023), introducing novel frameworks for comprehensive evaluation of generative systems.
• Analyzed real-world ethical and governance implications of AI deployment, focusing on bias mitigation, resource allocation fairness, and policy robustness.

Systems Analysis & QA Intern - TechNova Solutions - Nairobi, Kenya

(2021-01 - 2021-05)

• Built automated test scripts that detected workflow anomalies, reducing manual review time by 30%.
• Collaborated with engineering teams to refine QA protocols and improve traceability between requirements and test cases.

Master of Science - Data Science and Artificial Intelligence - Carnegie Mellon University (2022-09 - 2023-05)

Bachelor of Science - Computer Science - University of Nairobi (2021-06)

Certificate - AI Ethics and Prompt Engineering - Coursera (DeepLearning.AI) (2024-08)