Job Description

Most interfaces are designed months before anyone uses them. We're building the ones that aren't — interfaces composed on demand, in response to what someone asked for, or what they're about to need.

The biggest decisions are still ahead: what the frameworks look like, where the runtime boundary sits, how we tell a good generated interface from a bad one. You'd be making them — working out how to meet the craft and polish people expect from Apple without giving up the range and surprise that make generative UI exciting.

As one of the first engineers on a newly formed team, you will partner directly with Apple's design studio and AI teams to invent this paradigm. You'll help set the technical direction, shape the team's culture, and establish the bar for our engineering.

One half of the job is interface craft: layout, motion, responsiveness, and the thousand details that make something feel native. The other half is building the system that decides at runtime what the interface should even be — generating something a designer would normally draw.

The work in front of us: productizing the execution runtime and hardening the security boundary around it; driving generation quality through prompt iteration, automated evaluation, and output validation; designing the UI frameworks that generated interfaces are built from; and shipping the features that people rely on. This requires a rare balance: rapid, highly ambiguous design exploration to find what feels right, followed by the rigorous engineering required to ship cross-platform frameworks at Apple scale. You will partner with AI/ML teams, app teams, framework owners, security engineers, and designers across Apple to take features from idea to shipping.

Preferred Qualifications

Because this is a foundational role on a new team, we don't expect you to have every skill listed below. We are looking for deep expertise in a few areas, and the engineering fundamentals to learn the rest:

Experience productizing a language runtime, interpreter, or VM under real workloads
Experience designing a DSL, schema, or API surface that another system or model generates against
Familiarity with sandboxing untrusted code and mediating its access to data and system capabilities
Ability to reason adversarially about an API surface
Experience defending against prompt injection — guardrails, trust boundaries for retrieved content, personal data in context
Experience evaluating nondeterministic systems: golden sets, automated judges, metrics that track what users notice
Hands-on experience with prompt design, tool definitions, structured output, and output validation
Experience shipping a production ML-backed feature at scale
Strong design sensibility, and the judgment to tell when a generated interface isn't good enough

Minimum Qualifications

6+ years of software engineering experience
Bachelor's degree in Computer Science or equivalent experience
Expert-level Swift, with production SwiftUI (or comparable declarative UI)
Led an application or framework end to end, including its public API and its adopters
Hands-on experience building user-facing features with generative models (shipped or prototyped)
Experience setting technical direction across teams and mentoring engineers