Machine Learning Engineer

Proton, the privacy-first tech company, is expanding its engineering team in Taipei with a full-time, onsite Machine Learning Engineer role. The company has grown from a small startup launched in 2014 by CERN scientists to a profitable, independent player behind Proton Mail and a family of privacy tools including Proton VPN, Proton Drive, Proton Pass, and Proton Calendar. With more than 500 team members across 50+ countries, Proton moves fast, keeps hierarchies light, and prioritizes meaningful impact over optics. They’re looking for curious, collaborative engineers who want to shape the future of secure AI.

The role sits within the MSA team, short for Mail Delivery, Spam, and Anti-Abuse, a multidisciplinary group formed in 2019 to tackle complex security challenges across the Proton ecosystem. The team builds sophisticated systems from the ground up, blending human insight with machine learning to make tens of millions of decisions in real time or asynchronously. Their focus areas include reliable mail delivery and spam prevention, abuse detection, account security, and site reliability and resilience. In recent years, their work has dramatically reduced spam misclassifications, blocked millions of abusive signups, protected hundreds of thousands of users, and mitigated hundreds of DDoS attacks. Proton has expanded impact with Proton Sentinel, an AI-human hybrid security program, and Proton CAPTCHA, plus they’re scaling capabilities with agentic AI for autonomous threat detection and responses. The company has grown from 2 engineers to 40+ engineers and analysts across 3 continents, operating around the clock, and offers a global, mission-driven environment where top talent thrives.

If you want to do work that matters at scale and collaborate with experts in security, ML, and systems engineering, this could be the fit you’ve been aiming for. You’ll join a team that values concrete outcomes, practical privacy, and bold experimentation, all within a company that lets you own meaningful pieces of the product from day one.

Responsibilities for Proton’s ML Engineer Role in the MSA Team

You will design and deploy scalable ML systems using modern MLOps practices, supporting real-time inference, model monitoring, and automated retraining pipelines. You’ll architect agentic AI systems that can autonomously detect threats, enforce policies, and initiate self-healing security responses. You’ll build tools for model development, debugging, and explainability to speed iteration and transparency, and you’ll optimize ML workflows for distributed environments to handle large-scale data processing. You’ll work closely with security analysts, backend and frontend engineers, and customer support to ensure technical solutions align with user needs. You’ll push forward ML methods for anti-abuse and account security, exploring techniques like adversarial robustness, graph-based anomaly detection, and privacy-preserving AI.

Who Should Apply for Proton’s ML Engineer Position

• Degree in Computer Science or a related quantitative field, with 2+ years hands-on experience in building and running ML systems.
• Production-level ML expertise: you’ve shipped models that handle real-world scale and complexity.
• Hands-on experience with LLMs in agentic environments.
• Strong software engineering skills: proficiency in Python, with a solid understanding of backend and server fundamentals.
• Statistical rigor: deep understanding of probability, hypothesis testing, and experimental design.

• Experience with adversarial ML or red-teaming ML systems.
• Experience working on distributed systems.
• Contributions to open-source security/privacy tools.
• Knowledge of federated learning, differential privacy, or on-device inference.
• Familiarity with graph-based anomaly detection or knowledge graphs.
• Technical leadership: ownership of projects, mentoring peers, and shaping technical direction.
• Security mindset: familiarity with abuse patterns, adversarial attacks, or threat intelligence (bonus points).

Compensation and Working Arrangements for Taipei Onsite Role

The position is onsite in Taipei, with Proton offering a comprehensive set of benefits designed to support your wellbeing and growth. The company emphasizes impact at scale, the opportunity to work with a diverse, high-caliber team, and access to the latest hardware and software to do your best work. Learning and development are a priority, with a culture that rewards ownership and real-world results. Employees receive strong health coverage, retirement options, generous leave, and wellness support, plus stock options from day one so you can share in the company’s success. In-person collaboration is valued, with offices around the world in locations such as Geneva, Zürich, Barcelona, and London, while flexible working hours are encouraged when outcomes are delivered. Daily meals and snacks are provided in offices, and transport allowances, public transport, bike schemes, or parking, are available. The company highlights that work should matter and that you’ll be joining a globally distributed, mission-driven, privacy-focused team. If you need adjustments during the hiring process, Proton’s talent partners can help, and privacy practices apply to how applications are handled.

Tips for Your Proton ML Engineer Application

Put your most relevant production ML work up front on your resume, especially any systems you built that run at scale and involve LLMs or agentic AI in security contexts.

Show concrete evidence of your must-have skills by linking to projects or code samples where you designed, deployed, and monitored ML models, and where you used Python in backend or server contexts.

Be ready to discuss topics central to the role, such as adversarial robustness, graph-based anomaly detection, privacy-preserving AI, and building ML pipelines for distributed environments, with specific examples from your experience.

Ask a targeted question about Proton’s approach to agentic AI and privacy, for example: how autonomous threat detection decisions are balanced with user privacy constraints in production systems and how success is measured in the MSA team's security outcomes.