Data Scientist · Machine Learning Engineer
I design, build, and deploy machine learning solutions — from computer vision for public health to RL research and climate-focused analytics. I care about clarity, reliability, and measurable impact.
01
I’m a Data Scientist and Machine Learning Engineer focusing on computer vision, sequence modeling, and MLOps. I care about reproducibility, clean UX for data tools, and delivering models that actually make decisions better.
I design ML workflows end‑to‑end: framing the problem, building robust datasets, prototyping, evaluating with the right metrics, and shipping simple interfaces teams can use.
Clear specs, tight feedback loops, and reproducible code. I prefer small, testable steps and dashboards that make model behavior obvious.
02
Research recognized through competitive awards, scientific presentations, and graduate funding.
August 2025 · Montréal
2nd Francophone Interfaculty Research Conference in Biostatistics (CFIRB), held at the CHUM Research Centre.
“Visualization Is All You Need: Robust Approach for Unsupervised Dimensionality Reduction in Genetics and Genealogy.”
Read the announcement ↗August 2025 · Montréal
Canadian Society for Epidemiology and Biostatistics (CSEB) Conference 2025.
View conference abstracts ↗2025–2026
Fondation Armand‑Frappier scholarship awardee in Prof. Amadou Diogo Barry’s laboratory, Experimental Health Sciences.
View awardees ↗2023 · Tübingen, Germany
Selected for the AIMS–University of Tübingen fellowship program. Research focused on randomized strategies for non-stationary reinforcement learning.
View fellowship profile ↗2024–present · Canada
Full funding for the Master’s degree in Biotechnology at INRS–Armand‑Frappier Santé Biotechnologie.
March–April 2024 · Rwanda
Fully funded Certificate in Data Science at the AIMS Research and Innovation Centre in Kigali.
2022–2023 · Senegal
Fully funded African Master’s in Machine Intelligence (AMMI), an intensive graduate research program in artificial intelligence.
July–December 2020 · Germany
Fully funded Certificate in Management and Technology at ESMT Berlin through a partnership with DAAD and AIMS South Africa.
2019–2020 · Rwanda
Full scholarship for the Master’s in Mathematical Sciences at AIMS Rwanda, specializing in Operator Theory and Quantum Computing.
03
Python · PyTorch · TensorFlow · Scikit‑learn · SQL · NoSQL · Docker · Linux
CNNs, UNet, transfer learning, augmentation, evaluation.
Recurrent RL, transfer/meta‑learning, bandits.
Feature engineering, rolling CV, deployment patterns.
Data versioning, CI, simple UX for model use.
04
Aix-en-Provence, France · Remote
Applied artificial intelligence to reduce greenhouse-gas emissions in freight transportation. Found shortest feasible routes, integrated ISO 14083:2023 into an existing data solution, developed multi-class segmentation and polygonization for satellite imagery, and improved emissions-estimation accuracy.
Python · TensorFlow · Keras · AWS · SQL/NoSQL · Object Segmentation · Time Series · GLEC · ISO 14083
Germany
Conducted research on transfer learning in recurrent reinforcement-learning problems. Designed new recurrent RL algorithms and analyzed experimental results for sequential decision-making, meta-learning, and bandit settings.
Funded by the Cluster of Excellence – Machine Learning for Science
Madagascar
Identified business challenges, designed data pipelines and experiments, developed predictive models, cleaned large datasets, presented complex results, and delivered data-driven insights.
Python · SQL Server · TensorFlow · PyTorch · Scikit-learn · R
Kigali, Rwanda
Supported lecturers, conducted tutorials, marked assignments and quizzes, assisted with academic quality assurance, supported IT and library operations, and completed administrative tasks.
05
Bioinformatics & AI with applications in health and spatial omics.
Foundational Methods in Data Science at the AIMS Research and Innovation Centre.
Deep learning fundamentals and research training.
Six-month intensive training across business technology and management.
Optimization, probability, and numerical methods.
Specialization in Probability and Analysis.
Foundations in mathematics, programming, and algorithms.
06
Complete work across dimensionality reduction, spatial omics, computer vision, reinforcement learning, climate analytics, bioinformatics, and data engineering.
Featured research · CosMAP
I developed CosMAP (Contrastive Manifold Approximation and Projection) which is a new graph-based unsupervised dimensionality-reduction method for producing faithful and interpretable embeddings of sparse, noisy data.
On MNIST, shown here, CosMAP organizes 70,000 handwritten-digit samples with 784 features into a two-dimensional map. Each point is one image; color indicates its true digit label. Compact neighborhoods and distinct regions make both local relationships and the overall class structure easier to inspect.
Result: On the MNIST and USPS benchmarks, CosMAP yielded the most interpretable two-dimensional cluster structure among the evaluated methods, with LocalMAP providing the closest competing performance. Its embeddings produced compact, well-separated digit populations while maintaining a readable global organization.
Comparative evaluation: CosMAP was evaluated against PaCMAP, UMAP, LocalMAP, t‑SNE, TriMAP, PHATE, PCA, h‑NNE, InfoNCE‑t‑SNE, Neg‑t‑SNE, and NCVis across diverse image, single-cell RNA-seq, and genealogical datasets.
The method combines cosine-similarity neighborhoods with temperature-normalized contrastive affinities, an attractive–repulsive objective, and a two-phase refinement strategy.
07
You can view or download my résumé below. For the most up-to-date version, please contact me.
Download PDF08
I'm open to collaborations, internships, and full-time roles in Data Science and Machine Learning.