Jaciana Nogueira

Validate LLM-generated responses by verifying factual accuracy through cross-referenced sources, extract supporting evidence and URLs, and enforce zero-hallucination policies with 95%–100% accuracy.

Evaluate AI-generated images for visual coherence, anatomy, artifact detection and prompt alignment, completing up to 60 evaluations/day with sub-2-hour turnaround and high precision.

Validated over 1000 AI model responses for accuracy, ethical coherence, and cultural appropriateness, maintaining 95%+ quality scores across multiple AI training platforms. Specialized in multilingual evaluation (English/Portuguese) and handled confidential data under strict security protocols.

Founded an AI education and research project producing accessible AI content, prompt frameworks, datasets and repositories while conducting independent research on LLM behavior, memory architectures, and safety.

Built comprehensive operational manuals and SOPs for a franchise system by structuring workflows and mapping processes to improve training and operational consistency.

Applied data analytics to enhance operational efficiency and conducted root cause analysis for system performance issues. Implemented Lean methodologies for process improvement, collaborating with cross-functional teams to achieve measurable results.

Postgraduate program focused on information security, cybersecurity, and applied artificial intelligence with emphasis on AI security and risk.

Gained international experience, enhancing a global AI perspective and advanced English proficiency, critical for international AI training platforms. Developed a multidisciplinary approach to technology challenges.

Developed skills in systems thinking, process optimization, data analysis, and problem-solving methodologies. This foundation is relevant for understanding and improving AI training workflows.