ZOUZOUA AXEL ISRAEL BLÉ

– Predictive Risk Modeling & Failure Forecasting: Developed supervised ML models on multi-year multivariate time-series data to predict failures in critical infrastructure. Achieved a 20% reduction in operational downtime, directly mitigating financial exposure and operational risk through early detection. – Real-Time Anomaly Detection Systems: Engineered an end-to-end pipeline (ETL → unsupervised learning → Streamlit dashboard) for continuous monitoring of high-value assets. The system flags deviations in real-time, enabling immediate intervention to prevent costly incidents and reduce unexpected losses. – Quantitative Forecasting for Resource Optimization: Built and productionized neural network models (MLP) for high-stakes environmental forecasting, enabling data-driven resource allocation and risk planning. Demonstrated impact led to client contract renewal. – Simulation & Scenario Analysis Tools: Migrated legacy numerical simulation models (MATLAB/Simulink) to a modular Python framework, deploying them as interactive applications for real-time what-if analysis and strategic decision support. – Al-Augmented Data Science Workflow: Integrated LLM-based tools to automate KPI extraction and feature engineering, accelerating model development cycles by ~30% and enhancing team productivity on data-intensive projects. – Production MLOps & Reliability Engineering: Established robust ML pipelines with automated monitoring, alerting, and retraining, improving system reliability by 30% and ensuring consistent performance of mission-critical models. – Technical Solution Architecture: Expanded role to include client-facing technical consulting, including incident triage, requirements analysis, and solution design. Acted as a bridge between client stakeholders and engineering teams to deliver robust, tailored solutions. – Product Ownership & Pre-Sale Technical Consulting: Delivered software and data service demonstrations to national and international stakeholders (technical and non-technical). Gathered client requirements, translated them into solution specifications with the engineering team, and presented resulting architectures directly to prospects.

– Research in Low-Latency Sequential Models: Developed deep learning architectures for real-time classification of high-frequency data streams, with a focus on adaptive inference and minimal decision latency. – Generative Modeling for Synthetic Data: Implemented and evaluated Generative Adversarial Networks (GANs) to address data limitations, building expertise in synthetic data generation for model robustness and scenario simulation. – Published Research Contribution: Authored a master's thesis on applied generative models, documenting the end-to-end research process from methodological design to experimental validation.

– Tutored students in programming fundamentals (Python, C++), data structures, and algorithmic problem-solving – Developed pedagogical materials and tailored learning paths for technical skill development