Principal/Senior Machine Learning Scientist - Search and Discovery

About the Role

With Roblox’s daily active users growing at a record pace, we are seeking experienced machine learning researchers who thrive on solving complex challenges and designing scalable, groundbreaking solutions. In this role, you will innovate on LLMs, VLMs, autoregressive models, or other large models for user understanding, content understanding, recommendation, and agentic search. Your work will lay the foundation for personalizing user experiences, enhancing engagement and retention, and driving growth within the creator ecosystem.

In search and discovery, you will contribute to the core systems that enable users to discover and interact with experiences across Roblox, covering query suggestions, autocomplete, retrieval, and ranking.

Responsibilities

• Design and implement machine learning solutions that power agentic search and discovery systems to drive Roblox content distribution.
• Design and implement large-scale recommendation models.
• Innovate on LLMs, VLMs, autoregressive models, or other large models for query understanding, user understanding, content understanding, recommendation, and agentic search.
• Design eval and post-training loops (SFT, preference optimization, and RL) for ranking, query understanding, and agentic systems.
• Balance researching new technologies with a practical approach to integrate research efforts into Roblox products.
• Collaborate with other teams across Roblox to build products with a community-first approach.
• Communicate with the industry and communities where applicable and act as an ambassador for Roblox.

Qualifications

• 4+ years (Senior) or 7+ years (Principal) of experience designing and building scalable machine learning solutions.
• Experience with transformer-based model training, inference, and product integration.
• Ability to break down product requirements into iterative deliverable stages.
• Demonstrated ability to initiate and facilitate challenging technical decisions across multiple teams and organizations.
• User-centric mindset with a creative and strategic problem-solving approach.
• BS, MS, or PhD in Computer Science or a related discipline, or equivalent experience.

Preferred Qualifications

• In-depth experience in recommendation, search retrieval and ranking, or LLM post-training and agentic systems.