Open to opportunities

alex gichangiru

@alexgichangiru

AI quality and data annotation specialist improving LLM reliability through red-teaming, JSON-structured training datasets, and conversation design.

Kenya

Message

What I'm looking for

I’m looking for full-time remote work helping train and evaluate LLMs—designing multi-turn, tool-based scenarios, running QA/red-teaming, and building high-quality JSON datasets that strengthen model robustness.

I’m a results-driven AI Training & Data Annotation Specialist focused on improving LLM robustness through quality evaluation, red-teaming, and multi-turn conversation design. At Mercor, I evaluate 500+ AI-generated responses with 95%+ accuracy and run red-teaming exercises to uncover 150+ edge cases.

I structure high-quality, tool-based assistant interactions and deliver consistent, QA-driven outputs using JSON formatting and function-calling workflows. As a freelance AI Annotation & Data Specialist, I annotate NLP and computer vision datasets, design realistic user–assistant conversations, and apply guidelines and QA checks to maintain dataset consistency—alongside projects covering function-calling simulations and YOLO/COCO-ready computer vision annotations.

Experience

Work history, roles, and key accomplishments

Current

AI Annotation & Data Specialist

Current

Freelance

Jan 2024 - Present (2 years 4 months)

Annotated NLP and computer vision datasets by designing realistic user–assistant conversations and tool-based workflows. Improved dataset quality through QA checks while maintaining consistency with provided guidelines.

Data Annotation Tool Based Workflow Modeling Guideline Based Consistency