02 / Experience

What I've
built.

Five roles across applied ML, healthcare imaging, research, and analytics — with a featured medical-imaging engagement currently in clinical validation.

Featured engagement · 2025 — Present

Lead AI Developer
at Idoni.

Clinical Validation · King Faisal Hospital, Rwanda

AI + IoT platform for early detection and monitoring of adolescent idiopathic scoliosis. Currently in clinical validation with King Faisal Hospital, Rwanda.

idoni.vercel.app↗

Stack

PyTorchDenseNet121ResNet50Computer VisionIoTMLOps

What I delivered

  • Designed and trained a DenseNet121 + SVM hybrid classifier for scoliosis screening from spinal images, reaching 95% accuracy on validation data.
  • Built a ResNet50 backbone with a regression head for Cobb angle measurement, achieving 3.4° mean absolute error — within the clinically accepted inter-rater margin.
  • Curated and preprocessed the imaging dataset end-to-end: labelling, augmentation, normalisation, train/val/test splits, and quality checks.
  • Integrated an ESP32 + MPU-6050 IoT wearable streaming accelerometer and gyroscope data to track spinal progression between clinical visits.
  • Deployed the full stack on Vercel with a MongoDB Atlas backend; images are processed in memory.
  • Documented experiments, failure modes, and limitations in the Chapter 4–6 technical report used for clinical validation discussions.

Work history

Where I've worked.

Jun 2024 — Oct 2024

AI Developer

Eleka Inc. · Rwanda

  • Designed and implemented AI models to optimise business processes and enhance decision-making.
  • Collaborated with cross-functional teams to integrate AI solutions into scalable applications.
AI ModelsAlgorithmsScalable Systems
Jan 2025 — Apr 2025

Data Scientist Intern

Only Quality Data · Ghana · Remote

  • Developed a Heart Disease Prediction model using classical ML, improving diagnostic signal extraction from tabular clinical features.
  • Refined models iteratively with mentors, tuning hyperparameters and evaluating with stratified cross-validation.
Classical MLHealthcareCross-Validation
Programme

Data Scientist & ML Trainee

University of Tokyo · GCI Lab · Tokyo, Japan

  • Conducted EDA and preprocessing of structured and unstructured datasets for predictive modelling.
  • Built and fine-tuned deep learning architectures (CNNs, RNNs, LSTMs) for supervised and unsupervised tasks using PyTorch and TensorFlow.
  • Collaborated with researchers to design experiments and validate ML-driven hypotheses.
Deep LearningPyTorchResearch
Sep 2024 — Dec 2024

Data Analyst Intern

Only Quality Data · Ghana · Remote

  • Analysed customer churn and employee attrition using Power BI and DAX, producing actionable retention recommendations.
  • Designed interactive dashboards for budgeting and financial planning.
Power BIDAXAnalytics

Full resume

For the long-form, grab the PDF.

Download Resume↗

Next

About→
Back to work