The Arrival Experience is a defining moment in how users interact with the real world through Maps. As we make navigation more intelligent and context-aware, this experience becomes central to delivering precision and confidence; helping users reach the exact place they intend to go. Whether it’s a storefront, parking entrance, or doorstep, we are building the intelligence that powers this crucial interaction for hundreds of millions of Apple Maps users. As a Machine Learning Engineer on the Arrival Experience team, you’ll design and deploy models that blend satellite and aerial imagery, behavioral signals, and map metadata to deliver precise, intuitive destination guidance. You’ll also develop generative AI-based multi-agent systems that can dynamically reason, evaluate, and improve guidance quality at scale. This role offers the opportunity to solve highly complex, real-world problems at the intersection of geospatial intelligence, computer vision, and generative AI. At Apple, we build products that enrich people’s lives, and we believe even the smallest moments, like arriving confidently at the right place, are part of that promise. As part of the Arrival Experience team, your work will help turn complex geospatial systems into intuitive everyday magic for hundreds of millions of users.

DESCRIPTION

Are you passionate about building intelligent systems that seamlessly integrate various data modalities? As a Machine Learning Engineer on our Arrival Experience team, you will: - Build and deploy ML models that integrate imagery, spatial context, and user behavior to enhance arrival accuracy - Design and implement generative AI-based multi-agent systems that evaluate and improve arrival experiences - Work closely with product and engineering teams to bring model intelligence into user-facing features - Evaluate models using both automated metrics and real-world feedback, continuously iterating for quality - Stay informed on advances in computer vision, generative AI, and applied machine learning to explore new approaches and keep the system adaptive and future-ready

MINIMUM QUALIFICATIONS

• MS or PhD in Computer Science, Machine Learning, or a related field.
• 2+ years of hands-on experience in machine learning, particularly in computer vision or multimodal model development
• Experience building and deploying transformer-based architectures, including Vision Transformers (ViT), or multimodal vision-language models such as CLIP
• Familiarity with multi-agent system design and orchestration frameworks, including the use of Model Context Protocol (MCP), A2A (agent-to-agent) communication systems, or LangGraph for generative agent workflows
• Proficiency in Python with deep expertise in PyTorch for model development and deployment
• Experience working with large-scale datasets and building scalable training or inference pipelines
• A self-starter with a go-getter attitude, able to work around data limitations or organizational constraints and continue making progress
• Excellent problem-solving skills with the ability to approach ambiguous problems creatively
• Strong collaboration and communication skills with a track record of working effectively across teams
• Commitment to fostering an open, inclusive, and team-oriented work environment

PREFERRED QUALIFICATIONS

• Demonstrated success applying generative AI to production systems beyond prompt engineering
• Familiarity with extracting and refining polygonal features from imagery, including post-processing techniques such as the Marching Squares algorithm or similar contour extraction methods
• Experience working with geospatial data formats such as GeoTIFF, raster maps, shapefiles, or GeoJSON

Please refer the Job description for details