Abderrahmane Pazraq

As Head of Engineering, I managed two cross-functional teams (totaling 10 engineers across Cloud and Machine Learning development), reporting directly to the CEO. I held full responsibility for implementing the go-to-market strategy and leading the technical engineering efforts for the company's cloud-based traffic light management platform, designed to compute optimal traffic light timings adapting to live traffic conditions. I supervised the end-to-end development lifecycle of state-of-the-art Multi-Agent Reinforcement Learning (MARL) agents for traffic light control, achieving up to a 30% reduction in vehicle wait time at intersections and a 10% reduction in CO2 emissions compared to traditional systems.

I also contributed to the design of the Statement of Work for a two-year government contract with a U.S. State Department of Transportation, overseeing the timely delivery of project milestones to clients. This role required a blend of leadership and managerial skills (team building, product management, strategic thinking) and technical expertise (system design, microservices architecture, Scala, Python).

At Microsoft, I worked as a Data Scientist Intern in the Office of the CTO as part of my end-of-studies internship for my master’s degree at École Polytechnique and HEC. My main project involved designing an end-to-end Responsible AI Lifecycle tailored for cloud-native machine learning applications. As part of this, I authored two white papers on Responsible AI that were shared with Microsoft employees worldwide. The role required expertise in privacy-preserving machine learning, cloud computing (Azure), and AI governance.

As Lead Data Scientist, I guided the Reinforcement Learning team in developing a Deep Q-Learning algorithm to optimize traffic light timings based on real-time traffic data. My responsibilities included overseeing the deployment and operation of the algorithm on cloud platforms, using tools such as Google Cloud Platform (GCP) and implementing MLOps practices, including CI/CD pipelines for machine learning models. This role required strong skills in reinforcement learning, Python, and cloud computing, as well as the ability to lead a technical team to deliver impactful, data-driven solutions for traffic management.