The future of AI is on-device. On the Siri team, we're solving the defining challenge: building and shipping powerful, LLMs that run with extreme efficiency across hundreds of millions of iPhones, Watches, and Macs. You will own the full ML stack, from data to deployment, bridging the gap between bleeding-edge research and product-critical models. This is where the hardest problems in applied ML are being solved.

Training & Fine-Tuning: Architect and train LLMs specifically for assistant behavior. This is your opportunity to define what "reasoning" and "task completion" mean for a billion users.

Performance & Optimization: Master the art of model surgery. You'll design and implement novel context and adapter strategies to make our models faster and smarter within the tight constraints of on-device deployment.

Data as a Product: You understand that world-class models are built on world-class data. You will own the curation, augmentation, and evaluation pipelines that are the lifeblood of our systems.

System Architecture: Your insights will shape not just the model, but the entire system. You'll work with cross-functional experts in software, research, and product to build a cohesive, intelligent experience.