Imama Muhammad

I’m an AI-focused data scientist with an MSc in Artificial Intelligence from Queen’s University Belfast, graduating with Distinction. I bring a research mindset to production problems—turning experiments, uncertainty, and model performance into real business outcomes.

Recently, as a Cyber-AI Researcher, I built optimized Graph Neural Network architectures inside the Nvidia Morpheus pipeline for intrusion detection on CIC-2018 and UNSW datasets. I used self-supervised Autoencoder approaches to achieve satisfactory accuracy.

At AFINITI, I progressed from Analyst Software Engineer to Data Scientist II and then Senior Data Scientist. I built and deployed production machine learning models with Bayesian frameworks (Stan, PyMC), Random Forest, and XGBoost—delivering $75K+ monthly incremental revenue and 5% average client gains, alongside churn prediction achieving 89% AUC with a 15% lift over baseline.

I also emphasize robust engineering and measurable experimentation: I engineered 200+ features and reduced to 45 high-impact predictors, automated monitoring pipelines for 99.5% uptime, and designed and analyzed 50+ A/B experiments using scipy and statsmodels. I built Bayesian experimentation frameworks, created Power BI dashboards for senior leadership and C-suite, and led 4 data scientists while mentoring junior team members.