Derek Yendoh
I build AI systems that work in the real world — from model training to clinical deployment.
I'm a software engineer and machine learning practitioner from Accra, Ghana. My work sits at the intersection of AI research and real-world systems — building tools that don't just predict, but actually help people make decisions.
My most significant project to date is DiaTrack, a clinical decision support system for Type 2 Diabetes risk prediction. I built it end-to-end: the ML pipeline, the FastAPI backend, the clinician dashboard, and the patient-facing mobile app. It's the kind of problem that doesn't fit neatly into a Kaggle competition — it required thinking about doctors, patients, data privacy, and deployment all at once.
I'm especially interested in AI applications in healthcare and education — domains where the gap between what's technically possible and what's actually deployed is still very wide. I want to help close that gap.
“The gap between what AI can do and what actually helps people — that's where I want to work.”
Real systems built for real problems — from model to deployment.
An AI system that helps clinicians assess diabetes risk — and helps patients understand their own health — built for real clinical use.
AUC: 0.978 · Accuracy: 98%
Read more →
A pipeline that automatically enriches AI model documentation from the HuggingFace Hub — making models easier to discover and evaluate.
Read more →
A production-grade framework for teams — shared standards, reusable templates, and department-specific prompt libraries for consistent AI workflows.
Read more →
Self-supervised learning pipeline that estimates cardiometabolic risk from photoplethysmography (PPG) signals captured on wearables.
Read more →
Tools grouped by what they do — not vanity percentages.
The AI and prediction layer
The infrastructure that makes it run
The interfaces people interact with
Education and community.
Kwame Nkrumah University of Science and Technology
Biomedical engineering foundation with focus on medical device design and health informatics.
Ghana NLP Community
Delivered a formal talk on fine-tuning on-device speech models for Ghanaian English; ongoing project sharing and technical contributions to community research initiatives.
Ghana Engineering Students Association (GESA), KNUST
Organised and facilitated tutorials on complex engineering subjects; mentored fellow students through academic challenges and project development. Volunteered as academic mentor to first-year computer engineering students for three consecutive years providing orientation, guidance, and ongoing academic support.
Open to collaborations in HealthTech, EdTech, and applied ML — and always happy to talk to fellow builders.