Machine Learning Engineering Intern, Personal Agentic Search

At Apple, we build products that become part of people's lives. Every day, people rely on their devices to hold their messages, documents, photos, and notes. The Personal Agentic Search team uses Apple Intelligence to help them find and act on that information in natural, intuitive ways.

We are engineers and researchers building AI-powered agents that understand a person's context, reason over their personal data, and retrieve exactly what they need, all while preserving the privacy Apple users expect. As an intern, you will help build and evaluate the machine learning systems behind these experiences and see your work shape features that millions of people could use.

This internship sits at the intersection of NLP, large language models, and information retrieval. You will work on real projects alongside experienced engineers and researchers, helping turn research ideas into robust, user-facing experiences. We are looking for someone with a strong foundation in computer science and machine learning, and a real curiosity about how language models and search systems work. You don't need to be an expert in every area, but you should be excited about applying ML to hard, user-centered problems.

Design, prototype, and evaluate ML-driven approaches to personal search and retrieval, using LLMs, embeddings, and on-device machine learning. Explore agentic workflows in which LLMs plan, call tools, and reason over multi-step search tasks. Build and improve NLP components such as query understanding, semantic retrieval, ranking, and re-ranking. Run experiments, analyze results, and develop metrics and datasets to measure search quality and model performance. Collaborate closely with AI/ML researchers and senior engineers to turn research ideas into robust, user-facing experiences. Write clean, efficient, and well-tested code in languages like Python, Swift, and C++. Present your findings and learnings to the team.

Currently pursuing a BS, MS, or PhD in Computer Science, Machine Learning, or a related field. Strong foundation in computer science fundamentals (data structures, algorithms, software design) and proficiency in at least one programming language such as Python, Swift, C++, or Java. Solid foundation in machine learning and working knowledge of NLP fundamentals (tokenization, embeddings, language modeling, text classification), gained through coursework, research, personal projects, or professional experience. Strong communication and collaboration skills, with a desire to work in a team environment.

Familiarity with LLM architectures (e.g., Transformers, attention mechanisms) and how they are trained and applied. Familiarity with search and retrieval algorithms, such as indexing, ranking, nearest-neighbor search, and relevance evaluation. Experience with agentic LLM workflows, such as planning, tool calling, and multi-step reasoning. Experience with on-device machine learning or privacy-preserving ML approaches. Experience designing experiments, metrics, and datasets to evaluate model or search quality. Proficiency in more than one of Python, Swift, and C++.