Applied Machine Learning Engineer - Security

Job description

Apple's Security Engineering & Architecture organization is responsible for the security of all Apple products. Passionate about safeguarding users, we believe that the best defense requires a great offense. When it comes to securing more than a billion devices running the world's most sophisticated operating systems, that means finding vulnerabilities first.

Our mission is to discover, understand, and exploit vulnerabilities across all layers of Apple’s platforms, and we believe that ML techniques significantly enhance our ability to do so. We are seeking an Applied Machine Learning Engineer who will help us invent and deliver these new methods and techniques.

This position provides rare exposure to a full‑stack view of security along with direct access to expert knowledge, unique datasets, and cross‑domain experience. Your contributions will materially raise the security of products used by billions and strengthen Apple’s ability to defend against adversaries.

Can you make a difference on this scale? Join our extraordinary group of security researchers, tool developers, and machine learning experts, and help protect all Apple users.
In this role, you will integrate deeply with security research teams to understand the challenges of analyzing large, complex systems across Apple’s full stack, from custom silicon and microarchitectural elements to boot ROMs, firmware, kernels, system frameworks, web browser and user applications. You will design and develop AI‑enhanced systems using large language models, agentic workflows and machine learning approaches that complement other analysis methods such as fuzzing, static & dynamic analysis, and manual inspection. Your work will leverage raw data and expert behavior to create practical, scalable approaches to help researchers navigate vast codebases, reason about intricate attack surfaces, and identify subtle weaknesses that are challenging to detect manually. You will also collaborate regularly with security researchers to validate and challenge your innovations during real-world security evaluations, ensuring that your work will directly affect meaningful impact.
<h3>Minimum Qualifications</h3>Solid understanding of machine learning algorithms (supervised, unsupervised, and reinforcement learning) and evaluation methods
Hands-on experience with agentic frameworks such as LangChain, LlamaIndex, or AutoGen to build multi-step, tool-augmented agent workflows
Familiarity with MLOps practices including model versioning, CI/CD pipelines, and experiment tracking tools such as MLflow or similar
Strong software engineering and programming skills in languages like Python, C, C++, Swift, Objective-C
Collaborative and effective problem-solving and analytical skills
<h3>Preferred Qualifications</h3>Experience scaling LLM training and fine-tuning (pretraining, SFT, alignment)
Strong enthusiasm for security, especially offensive security
Familiarity with software‑analysis techniques such as fuzzing, static analysis, code‑analysis tooling, reverse engineering, binary-analysis
Collaborative and effective problem-solving and analytical skills