[2026] Senior Machine Learning Engineer, Multimodal AI, Computer Vision and Graphics - PhD Early Career

<div class="content-intro"><p><span style="font-weight: 400;">Every day, tens of millions of people come to Roblox to explore, create, play, learn, and connect with friends in 3D immersive digital experiences– all created by our global community of developers and creators. </span></p> <p><span style="font-weight: 400;">At Roblox, we’re building the tools and platform that empower our community to bring any experience that they can imagine to life. Our vision is to reimagine the way people come together, from anywhere in the world, and on any device.</span><strong> </strong><span style="font-weight: 400;">We’re on a mission to connect a billion people with optimism and civility, and looking for amazing talent to help us get there. </span></p> <p><span style="font-weight: 400;">A career at Roblox means you’ll be working to shape the future of human interaction, solving unique technical challenges at scale, and helping to create safer, more civil shared experiences for everyone.</span></p></div><p>At Roblox, computer vision and graphics power the way our global community discovers, creates, and interacts on our virtual platform. This role involves building cutting-edge models to interpret and analyze every form of content—experiences, text, images, videos, and 3D avatars. Your work will directly influence core systems that drive the <strong>next generation of </strong>creation, search, recommendations, and trust safety initiatives across our massive ecosystem.</p> <p>We’re looking for PhD candidates passionate about the intersection of computer vision, graphics, and generative modeling. You’ll work on applied research and engineering projects with direct production impact — enabling new ways for creators and players to bring ideas to life.</p> <h3><strong>Teams Hiring for This Role</strong></h3> <ul> <li><strong>Content Understanding:</strong> Develop large-scale computer vision and multimodal models that classify, organize, and recommend 3D and visual content to improve user discovery, personalization, and safety.</li> <li><strong>Account Identity</strong> : focuses on various aspects of user Identity from bot detection, alternate account detection graph, to age estimation based on behavioral and facial features. </li> </ul> <p><strong>You Will:</strong></p> <ul> <li>Lead the research and development of deep learning models for <strong>visual content understanding and 3D content generation</strong> (image/video/3D scenes, avatars, and assets).</li> <li>Design and implement foundation models for <strong>visual and 3D-based creation, search, and recommendations</strong>, ensuring a high level of fidelity, relevance, and ranking.</li> <li>Break down complex product requirements into iterative deliverable stages, moving applied research into high-scale production systems.</li> <li>Implement innovative visual and multi-modal models that power core Roblox functions (e.g., world creation, avatar systems, search, and recommendations).</li> <li>Build high precision facial age estimation across demographics from ground up including various fraud detection techniques for a robust and safe user identity system. </li> </ul> <p><strong>You Have</strong></p> <ul> <li>Possessing or pursuing a PhD in computer science, engineering, or a related field, with a thesis aligned to Roblox’s research areas.</li> <li>Expertise in one or more areas: computer vision, multimodal learning, 3D Graphics, or large-scale representation learning.</li> <li>Experience developing and training deep learning models using modern frameworks (PyTorch, TensorFlow, JAX).</li> <li>A strong research track record, evidenced by multiple publications and presentations in top-tier, peer-reviewed venues.</li> <li>Proficiency in one or more programming languages (e.g., Python, C++, Go, Java) and experience building and optimizing large-scale systems.</li> </ul> <p>Apply now to be considered for anticipated positions.</p> <p>You may redact age, date of birth, and dates of attendanc ... (truncated, view full listing at source)