Mohammed Shaban

• Lead AI/ML-based project delivery, articulating complex genomic and clinical insights to senior leadership and C-suite
• Developed and optimized ML workflows reducing delivery cycles and maximizing efficiency across business operations
• Accelerated 2 drug programs by 6 months through prioritized experiments and effective resource allocation
• Influenced $2M budget allocation for AI-driven drug discovery by presenting ROI-focused roadmaps to C-suite and board
• Applied SHAP-based explainability (80–85% interpretability) for actionable insights via management dashboards

• Built ensemble models (XGBoost, Random Forest) on 1,500+ structured datasets
• Drove model performance from 70% to 80% through iterative preprocessing, feature selection, and cross-validation
• Delivered analytical insights via dashboards and presentations to management and stakeholders

• Conducted EDA to identify patterns and anomalies using Pandas, Matplotlib, and Seaborn
• Documented technical workflows and provided insights to internal teams
• Supported junior talent development and onboarding processes

Integrated patient clinical and molecular data for 100+ patients using ensemble ML, achieving 0.82 AUC. Designed predictive algorithms validated across 3 oncology biomarkers (92% sensitivity, 89% specificity) to drive targeted interventions.

Refined 500+ patient sample data via Python and Bioconductor, reducing experimental time by 3 weeks and improving throughput.

Collaborated with multi-disciplinary teams to analyze 100+ compounds, achieving 65% improvement in binding affinity and reducing experimental testing time by 35%.

Engineered feature extraction pipeline processing 500GB+ of multi-omics data (RNA-seq, proteomics), reducing dimensionality by 95% while retaining biological signals.

Conducted retrospective clinical analysis (N=73) uncovering patient subgroups with 2.3x higher treatment efficacy.

Implemented predictive modeling on the Pima Indian Diabetes Dataset achieving 85% accuracy with a 90% training and 10% testing split.

Proposed a novel distance calculation formula reducing computation time by 25% and increasing nearest-neighbor accuracy by 15% in KNN.

• Biology of Cancer - Coursera, 2024
• Computational Data Science - Case Western Reserve University & INSOFE, 2023