Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I build systems that sit at the intersection of AI and the real world — full-stack applications, data pipelines, and human feedback loops that make AI models more useful, safe, and grounded.\n\nI am a MERN stack developer and AI data evaluator who does two things most engineers do separately: ship responsive web applications end-to-end, and work directly on the quality layer of AI systems—evaluating LLM outputs for safety, helpfulness, and factuality. I combine build and evaluation to close the gap between model output and real-world needs, enabling Forward Deployed, Applied AI, and Developer Advocate work that moves projects from concept to measurable impact.…I build systems that sit at the intersection of AI and the real world — full-stack applications, data pipelines, and human feedback loops that make AI models more useful, safe, and grounded.\n\nI am a MERN stack developer and AI data evaluator who does two things most engineers do separately: ship responsive web applications end-to-end, and work directly on the quality layer of AI systems—evaluating LLM outputs for safety, helpfulness, and factuality. I combine build and evaluation to close the gap between model output and real-world needs, enabling Forward Deployed, Applied AI, and Developer Advocate work that moves projects from concept to measurable impact.

Stephen Sifa Mwatsaka

AI Engineer, Full Stack Developer, Data Analyst, +1





I build systems that sit at the intersection of AI and the real world — full-stack applications, data pipelines, and human feedback loops that make AI models more useful, safe, and grounded.\n\nI am a MERN stack developer and AI data evaluator who does two things most engineers do separately: ship responsive web applications end-to-end, and work directly on the quality layer of AI systems—evaluating LLM outputs for safety, helpfulness, and factuality. I combine build and evaluation to close the gap between model output and real-world needs, enabling Forward Deployed, Applied AI, and Developer Advocate work that moves projects from concept to measurable impact.…I build systems that sit at the intersection of AI and the real world — full-stack applications, data pipelines, and human feedback loops that make AI models more useful, safe, and grounded.\n\nI am a MERN stack developer and AI data evaluator who does two things most engineers do separately: ship responsive web applications end-to-end, and work directly on the quality layer of AI systems—evaluating LLM outputs for safety, helpfulness, and factuality. I combine build and evaluation to close the gap between model output and real-world needs, enabling Forward Deployed, Applied AI, and Developer Advocate work that moves projects from concept to measurable impact.

Available to hire

I build systems that sit at the intersection of AI and the real world — full-stack applications, data pipelines, and human feedback loops that make AI models more useful, safe, and grounded.\n\nI am a MERN stack developer and AI data evaluator who does two things most engineers do separately: ship responsive web applications end-to-end, and work directly on the quality layer of AI systems—evaluating LLM outputs for safety, helpfulness, and factuality. I combine build and evaluation to close the gap between model output and real-world needs, enabling Forward Deployed, Applied AI, and Developer Advocate work that moves projects from concept to measurable impact.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Language

English

Advanced

Swahili

Fluent

Work Experience

AI Data Annotator & Evaluator at Minddrift

December 1, 2024 - June 1, 2025

Evaluated thousands of LLM response pairs to optimize safety, helpfulness, and factuality benchmarks; contributed to RLHF-style alignment workflows used in production AI systems. Assessed pairwise search engine results to improve ranking algorithms, and audited AI-generated content against product taxonomy standards for an e-commerce platform. Delivered structured, calibrated feedback to improve annotator consistency and output quality across the team. Operated across NLP evaluation, search quality assessment, and product categorization to build cross-domain understanding of AI system failures and fixes.

Full-Stack Web Developer (MERN) at Self-Employed

January 1, 2023 - Present

Designed, built, and deployed 5+ production-grade responsive web applications using React, Node.js, Express, and MongoDB — from requirements gathering to post-launch maintenance. Delivered all projects on time with 100% client retention, managing the full project lifecycle: scoping, development, testing, and deployment. Translated non-technical client requirements into functional specifications and clean, maintainable codebases. Built mobile-first, accessible UIs optimized for performance across devices.

Data Analyst at Self-Employed

May 1, 2020 - Present

Transcribed, cleaned, and structured audio and text datasets for AI training pipelines, improving data quality upstream of model development. Delivered analytical solutions across multiple concurrent client engagements, consistently meeting scope and deadline requirements. Applied structured problem-solving to convert raw, unstructured data into actionable insights aligned with client goals.

Youth Mentor & Technical Trainer at Compassion International Kenya – AIC Kilifi

May 1, 2020 - Present

Facilitated life skills and professional development sessions reaching 30+ youth per quarter; trained 15+ youth in technical and creative skills, breaking down complex concepts for non-technical audiences; supported community talent development programs contributing to measurable growth in youth engagement and skill acquisition.