Lead ML Engineer: Fraud & Anomaly Detection, Production

Ubisoft is natively data-driven: from the core of our games to our marketing actions, we’re leveraging data and algorithms.

Part of the Ubisoft Data Office, the role of the Consumer Data Domain is to leverage data to improve the consumer experience engaging them all along their journeys while making sure they feel safe and cared for.

Your role:

Reporting to the Director of the Consumer Domain, your role is to build scalable machine learning systems to preserve the trust and safety of our users. Our motto is to never stop at the prototyping phase by combining research and engineering cycles with the objective to deploy production-grade products.

In that context, we’re looking for a Lead Machine Learning Engineer with past practical experience, ideally in the field of anomaly / fraud detection.

Missions:

On the border between individual contribution and leadership, the archetypal missions for this position would be to:

• Take ownership over the projects you build and push them ahead.
• Evaluate new machine learning techniques and models, implementing them from scratch if needed.
• Write, optimize, and produce high quality code that can run at scale, using modern best practices (MLOps).
• In collaboration with data and software engineers, ship models or prediction pipelines to production in our internal or public cloud infrastructures.
• Lead and manage a team of 2 ML engineers focusing on detection topics.
• Develop and execute a comprehensive strategy to enhance the security of our users and reduce fraud.

QUALIFICATIONS

Whether you’re a software engineer who loves Machine Learning, or a research person that loves engineering, the key traits needed for this role are:

• Strong Python programming skills.
• Knowledge of Machine learning & Deep Learning frameworks.
• Experience with anomaly detection, unsupervised and/or semi-supervised learning is a strong plus.
• A “try hard, fail fast” mindset: research is evolving fast, and no one can know all the subfields of Machine Learning by heart. But we are never afraid to throw a punch at a new paper or an open-source repository with messy code, and so shouldn’t you be.
• Experience deploying models to production.
• Strong communication skills (English mandatory).
• Experience leading a team.