Job description

As engineer on this team, you will own the full lifecycle of our abuse detection machine learning models. You will collaborate closely with researchers to understand the threat landscape and partner with software and product teams to deploy robust, scalable defenses. We believe the most effective security systems are built by engineers who can translate adversarial insights into production-ready code. Your work will directly contribute to the architecture of Apple's AI platform and protect users from real-world harm. Here is what you will do: * Design, build, and deploy production-grade ML models to detect and mitigate abuse across multiple modalities (text, image, audio). * Own the full ML lifecycle: from prototyping and data analysis to deployment, monitoring, and the continuous improvement of models in production. * Drive the data strategy to continuously improve model performance by analyzing distribution gaps, contributing to synthetic data pipelines, and creating automated annotation systems. * Architect end-to-end systems for monitoring platform activity, detecting misuse, and triggering automated enforcement actions in real-time. * Collaborate with cross-functional partners in engineering, research, and product to define project requirements, establish technical direction, and deliver robust security solutions. 2+ years experience shipping machine learning models to production. You have owned the end-to-end lifecycle of a model, from development to deployment and maintenance. Strong familiarity with research fundamentals, machine learning principles, and development methodologies around LLMs, foundation models, and diffusion models Proficient programming skills in Python and deep learning toolkits (e.g. JAX, PyTorch, Tensorflow) Ability to work with sensitive and offensive content as part of building robust security and abuse detection systems. BS, MS or PhD in Computer Science, Machine Learning, or related fields or an equivalent qualification acquired through other avenues Hands-on experience with fine-tuning or aligning large language models for security or safety applications. Experience building large-scale data processing pipelines and ML infrastructure. Experience driving technical projects and collaborating with large, diverse, cross-functional teams.

Requirements

Must-haves

Design, build, and deploy production-grade ML models to detect and mitigate abuse across multiple modalities (text, image, audio).
Own the full ML lifecycle: from prototyping and data analysis to deployment, monitoring, and the continuous improvement of models in production.
Drive the data strategy to continuously improve model performance by analyzing distribution gaps, contributing to synthetic data pipelines, and creating automated annotation systems.
Architect end-to-end systems for monitoring platform activity, detecting misuse, and triggering automated enforcement actions in real-time.
Collaborate with cross-functional partners in engineering, research, and product to define project requirements, establish technical direction, and deliver robust security solutions.
2+ years experience shipping machine learning models to production. You have owned the end-to-end lifecycle of a model, from development to deployment and maintenance.
Strong familiarity with research fundamentals, machine learning principles, and development methodologies around LLMs, foundation models, and diffusion models
Proficient programming skills in Python and deep learning toolkits (e.g. JAX, PyTorch, Tensorflow)
Ability to work with sensitive and offensive content as part of building robust security and abuse detection systems.
BS, MS or PhD in Computer Science, Machine Learning, or related fields or an equivalent qualification acquired through other avenues
Hands-on experience with fine-tuning or aligning large language models for security or safety applications.
Experience building large-scale data processing pipelines and ML infrastructure.
Experience driving technical projects and collaborating with large, diverse, cross-functional teams.

Nice-to-haves

None flagged.

Application prep

Use the role’s language and application details while tailoring your materials.

Keywords to mirror

machine learningaimlpython

Platform: apple
Effort: medium