ML Engineer Creator Studio

Creator Studio is building agentic intelligence inside Apples professional creative apps. We aim for machine intelligence to assist and enhance creators intent and vision complementing creators rather than replacing it. This team has unusual leverage: our work spans many apps and getting a capability right lifts a whole family of professional creative tools at once. nnThe team is established and growing. Were looking for ML engineers who want to define the product. That means less time on models in isolation and more on the experiences they make possible working next to the designers quality teams app teams and engineers who own the workflows.

You work where a frontier model meets the real depth of a professional creative app. The job is to make that meeting useful. A models raw capability is not yet a feature and turning it into something that helps inside a real project and holds up well enough to ship is most of what youll do. Creative work is open-ended. That makes it hard to do well and hard to measure. So youll spend as much time on evaluation prompting and the tools and skills a model reaches for as on the models best creative software makes someone more capable while leaving them in charge. That balance is the hard part. The assistant should help a person move faster and push an idea further with the craft still theirs and getting it right is a product question as much as a modeling one. It sits at the center of this role.

Design build and evaluate agentic AI capabilities that let people accomplish real work in professional creative frontier model capability into reliable product behavior through agentic architecture tool and skill design and structured rigorous evaluations and quality measures for open-ended hard-to-score creative tasks and use them to drive model and system with on-device and server foundation models including optimizing for the constraints of on-device model and agent failures end to end from the user-visible symptom to the underlying with product engineers designers the teams behind our creative apps and creative-domain experts to translate real workflows into capabilities.

3 years developing and shipping reliable maintainable testable code or equivalent demonstrated -on experience applying LLMs / foundation models or multimodal ML to real products or projects e.g. prompting fine-tuning retrieval evaluation or agentic intuition for experimental design and a repertoire of statistical techniques to measure genuine effects from strong product sense: the judgment to ask not only can the model do this but is this the right experiencenAbility to work effectively in a fast-paced environment and to communicate clearly with technical and non-technical degree in Computer Science or a related field or equivalent practical experience.

Experience building agentic systems LLM tool-use / function-calling or evaluation hands-on experience as a professional or power user of creative tools DAWs NLEs image editors office productivity apps or design suites. You know these workflows from the inside and you have a point of view about what helps a creator and what gets in the way. Experience with Apples own creative apps is not welcome non-traditional backgrounds: demonstrated product judgment and creative-domain expertise can matter as much as a conventional Computer Science or ML programming skills in Python and experience with deep-learning toolkits like PyTorch JAX or optimizing models and algorithms to run efficiently on resource-constrained / on-device platforms such as Core with Swift and iOS/macOS record of publications or patents in relevant areas.