Applied Machine Learning Engineer - Security

Apple seeks a Mid-level Applied Machine Learning Engineer in Zürich to develop ML-driven tools for vulnerability discovery within its Security Engineering organization.

This position sits within Apple’s Security Engineering & Architecture group, which is tasked with protecting the security of every Apple product. The primary objective is to uncover, analyze, and exploit weaknesses across all platform layers by creating novel machine learning techniques that improve how vulnerabilities are found.

Responsibilities

• Partner with security research teams to grasp the difficulties of examining massive, intricate systems throughout the entire Apple stack.
• Build and refine machine learning systems that utilize large language models, generative models, and agentic workflows.
• Use raw data and expert patterns to build scalable solutions that assist researchers in navigating code and spotting flaws.
• Work closely with security researchers to test and validate new innovations during actual security assessments.

Requirements

• Deep knowledge of machine learning, with a specific focus on large language models and generative modeling.
• Strong interest or hands-on experience in security, particularly offensive security practices.
• Proficiency in software engineering using languages like C, C++, Python, Swift, Objective-C, or Rust.
• Strong analytical abilities and a collaborative approach to solving complex problems.

Nice to have

• Knowledge of software analysis methods such as fuzzing, static analysis, reverse engineering, and binary analysis.
• Understanding of security mitigations used in modern operating systems.

About the company

The team is driven by a passion for protecting users, operating on the principle that the most effective defense requires a strong offensive capability. They prioritize finding vulnerabilities early to ensure the security of over a billion devices worldwide.

• A strong commitment to user safety through proactive security measures.
• A belief that identifying weaknesses first is key to securing a massive global device base.