Zakaria Kaddache

Machine Learning Engineer | Freelance – BNP Paribas Cardif Datalab. – Design, develop, and deploy advanced Deep Learning models (NLP & Computer Vision) within production-grade MLOps pipelines. – Build and maintain CI/CD pipelines to accelerate ML workflows, ensuring reproducibility, scalability, and continuous delivery. – Promote best practices across the ML lifecycle, including model versioning, automated testing, and workflow automation.

As part of the Computer Vision team at Upciti, operating in the smart city sector, we leverage real-time edge-based computer vision to optimize urban services such as traffic management. Main activities: – Plan, execute, and analyze experiments to identify models and approaches that meet performance and quality objectives. – Package computer vision solutions and integrate them into production embedded devices, adhering to industry best practices. – Contribute to the design and development of internal tools for managing and streamlining data pipelines. Related activities: – Contribute to building and maintaining CI/CD pipelines, improving automation and deployment processes. – Supervise a 6-month end-of-study internship focused on a computer vision project, providing guidance and overseeing the intern's progress. – Participate in various team-related tasks, including code reviews, software design, and brainstorming sessions to foster collaboration and innovation.

At Coddity, an IT consulting company, my principal responsibility was to contribute to the Software Factory team of SFR (a French telecommunications company), where I developed data mining solutions to support decision-making. The Software Factory team at SFR develops and operates innovative products designed for all entities within the Altice Group, including B2C, B2B, RMC Sport Media, Next Advertising Agency, and Networks. Main activities: – Conduct experiments with fraud detection models and approaches applied to payments through direct carrier billing. – Develop and operate remote support software to track and optimize the management of technical support calls. – Conduct experiments to build telco-verified IDs to power internal marketing campaigns. Related activities: – Supervise and manage a team of interns to design and build a mobile application for a french city.

As part of the Amadeus Revenue Accounting team, at Amadeus IT Group operating in the travel industry sector, my mission was to design and develop a real-time anomaly detection solution for airline financial applications. Main activities: – Plan, execute, and analyze experiments of anomaly detections models and techniques for time series. – Develop a scalable and multi-user platform, to run anomaly detection jobs over real-time production data. – Empower the platform with monitoring dashboards and email alerting system to track financial automated workflows. Related activities: – Build the CI/CD pipelines for the anomaly detection platform. – Collaborate with department technical teams to collect relevant data and KPIs.

As part of the nuclear division energy (DEN), at the french alternative energies and atomic energy commission (CEA), my mission was to automate the configuration of a numerical simulator with nuclear applications. Main activities: – Functional and technical design of the software architecture. – Development, test and integration of the solution to the simulator.