Machine Learning Engineer

Machine Learning Engineer Location: Paris, France Department: Engineering About Keystone AI Keystone AI’s mission is to deliver the credential security every business and employee needs to thrive. Millions of consumers, and over 25,000 brands worldwide, such as Michelin, Air France, and Forrester, trust Keystone AI for industry-leading innovations, patented zero-knowledge security, and an unmatched user experience. Founded in Paris, Keystone AI has since established offices in New York and Lisbon, and has grown to more than 300 Keystone AIrs globally. We're looking for people who actively use AI tools to drive efficiency, creativity, and impact in their work. At Keystone AI, we drive innovation and value learning, strive for excellence in everything we do, and thrive as one team. Learn more about life at Keystone AI , including how we work , how we hire , and the benefits of being a Keystone AIr . About the role: You'll work on cutting-edge research in web understanding, using large labelled and unlabelled datasets of web pages. As a member of the ACE team, you'll help pioneer a state-of-the-art web analysis that will make life easier and more secured for millions of people worldwide. Its mission is to enhance user security and core product features through intelligent, AI-powered capabilities, such as improving the autofill experience, detecting credential risks, and protecting against phishing attacks. The team comprises Software Engineers, and Machine Learning Engineers, to ensure the quality of our analysis. We leverage advanced Machine Learning techniques to build intelligent security tools within a zero-knowledge architecture. Our blog post demonstrates how we implement autofill capabilities without accessing user data. We also explain our privacy-preserving approach to real-time threat detection and phishing prevention . We also advocate for public specifications for field autofill, like Semantically Annotated Web Forms . Join our team and help us make a difference in the world! Location: You will be based in Paris, with a hybrid policy (3 office days / week) and English as your working language. At Keystone AI, we embrace a hybrid culture that combines the best of both worlds: the creativity and energy of in-person collaboration with the flexibility of remote work. Our model is designed to strengthen team connections while supporting individual productivity and work-life balance. To maximize collaboration, we come together in the office on Mondays, Tuesdays, and Thursdays, while Wednesdays and Fridays offer more flexibility for focused work. About our stack: Languages: Backend: NodeJS, Typescript Frontend / internal tooling: React, Typescript ML: Python ML: Scikit-learn, Optuna, Onnx MLOps: Dvc, Mlflow Versioning: Gitlab CI/CD: Gitlab, Sonarqube, Terraform AI : Claude Code At Keystone AI you will: Work on R&D initiatives around web pages understanding and their applications to a Credential Protection ( password manager, identity provider, threat detection) Collect data, train model, make hypothesis, experiment and analyse results Deploy, evaluate, and continuously improve AI-powered features and agentic systems in production, including LLM-based applications, retrieval augmented generation (RAG), tool-using agents, and workflow automation Help maintain and improve our existing data pipelines (data collection tools and databases, model training pipelines, integrations into the Keystone AI web extension) Work with ML Engineers, Software Engineers, product managers, designers in a highly collaborative environment where everyone shares ownership of the company's success Act as an AI Pioneer within the engineering organisation, demonstrating proven experience integrating modern GenAI tooling into day-to-day workflows Learn every day, and share your knowledge with your coworkers Requirements: A Master / Engineer degree in Machine Learning/AI or Computer Science 4+ years of professional experience in product led company Agile team with in