AIML Sr Software Engineer AI, Evaluation

Do you want to help build the next generation of Apple AI products Our Evaluation organization measures quality across Apple Intelligence Siri and newest generative features and our measurements directly inform launch decisions. Our team builds the platforms behind it: LLM-as-judge autograders tools to validate them against human judgment and systems that run them at scale. nWe are looking for a senior engineer to own a critical part of that platform and take it to the next level.

This is an ownership role. You will take on major systems in our evaluation platform existing or new and be accountable for its direction quality and adoption. nThis role sits at the intersection of AI modeling software engineering and product quality. nWe are expanding rapidly so the ability to build a strong foundation and structure to any system is paramount. We are supporting features that encompass both product development and research which requires a balance of respect for both the scientific process as well as the product release lifecycle. nOur systems connect many platforms and serve teams across Apple so judgment with people matters as much as technical depth.

Own a core evaluation system end-to-end for multiple features: architecture roadmap reliability and developer testing and validation strategies that catch real failures before partners and maintain integrations with annotation compute reporting and benchmarking with teams building LLM judges on our platform turning their recurring needs into the bar for using AI coding agents without compromising -oriented: treats internal users needs as the job never loses sight of Apple customers and product quality and communicates clearly to align cross-functional partners.

BS/MS/PhD in Computer Science Machine Learning or a related field.n8 years of software engineering experience including owning production systems used by other Python skills and a strong engineering understanding of system design API design system testing and validation debugging and in shipping production code with AI coding agents while keeping quality high.

Proficient in building agentic systems LLM-based applications LLM-as-judge systems and offline to interpret evaluation results including human agreement run-to-run consistency and respecting constraints like data retention and privacy in building and maintaining technical integrations across multiple platforms in a large organization with varying technical requirements and needs based off scaling volume throughput -minded turning ambiguous requirements into a focused plan.