As a Freelance Data Scientist / Developer, I’ve evaluated machine learning models across 15 projects using accuracy, F1, and RMSE, improving average performance by 86%.
I’ve analyzed datasets of up to 2,500 records to surface trends and anomalies, sharing findings through dashboards and written reports. I also automated data cleaning and reporting with Python and SQL, cutting manual processing time by 90%.
For the Movie Recommendation System at ExploreAI Hackathon 2026 (Kaggle), I built a collaborative filtering model and reduced RMSE by 2.4% through Bayesian smoothing, ALS bias modeling, and ensemble blending.
I also built a Kenyan court-case NLP classifier that reached 85% accuracy and an 0.85 F1 score. I’m studying for a Bachelor of Science in Data Science at The Open University of Kenya, expected in 2028.
