Mondher Ilyes Izemrane

Achieved 72% model accuracy by training machine learning models on engineered UFC fighter statistics using scikit-learn and advanced feature selection pipelines

Built a secure, tiered-access interface with Next.js and Supabase, restricting AI predictions to authenticated users and enhancing engagement with interactive data visualization

Deployed cost-optimized containerized microservices with Docker and Vercel, enabling scalable updates without backend downtime

Developed an educational Generative AI product for 10,000+ children, leveraging reinforcement learning and cloud-based deployment

Reduced infrastructure and inference costs by 90% through cloud pipelines and model optimization

Facilitated stakeholder interviews and workshops with over 5 clients to gather technical and business requirements

Iteratively refined over 5 RFP drafts to ensure 100% alignment with evolving stakeholder goals, improving clarity and buy-in

Provided weekly 1:1 mentorship and academic strategy sessions to 12+ first-year students, leading to measurable GPA improvement and reduced dropout risk

Analyzed program feedback from 30+ participants to identify support gaps, contributing to updated coaching materials and improved session relevance

Increased student confidence by 50% using data-driven retention strategies, improving academic success

Built HeriTrace, an AI-powered ancestry tool that predicts surname origins with 85% accuracy and generates culturally themed images using Stable Diffusion

Implemented an interactive location map to capture user context and refine heritage predictions, improving relevance of results and user engagement

Developed a responsive React + Vite frontend, integrating with backend APIs to deliver seamless input flows and dynamic result displays

Designed and delivered technical workshops on BioPython and BLAST, creating custom datasets and code labs to teach bioinformatics workflows to 50+ students

Coordinated technical details with cross-functional teams to align coding events, speaker panels, and competitions with departmental schedules

Ensured smooth execution and strong student turnout for all technical programming initiatives

Led client discovery sessions to gather detailed requirements and iteratively deliver 3 customized web prototypes, achieving 90% satisfaction on final implementation

Improved search experience and data accuracy by implementing a dynamic live search bar, reducing redundancy and boosting frontend performance

Designed and normalized SQL database schemas to support efficient CRUD operations, streamlining meal option creation, editing, and updates