Principal Applied Scientist

Unity Technologies•San Francisco, CA

71d•$188,000 - $358,000

About The Position

At Unity, we’re shaping the future of real-time 3D development by enabling creators to harness intelligent, autonomous systems. As we expand our next-generation agentic AI—adaptive, goal-driven, and capable of reasoning about complex environments—we’re investing in research that pushes how large models perceive and understand 3D worlds. We’re looking for a Principal Applied Scientist to lead this work. In this role, you'll focus on advancing multimodal LLMs, strengthening spatial reasoning, and building world models that help creators author rich, interactive experiences. Your research will directly influence how agentic systems are built inside the Unity engine and how developers bring their ideas to life through more intuitive AI-driven workflows. This role is ideal for someone excited to explore new frontiers in multimodal modeling, grounded reasoning, and AI-assisted content creation.

Requirements

Advanced degree (MS or Ph.D.) in Computer Science, Machine Learning, Robotics, or a related field - or equivalent practical experience.
Deep expertise in multimodal LLMs, spatial reasoning, world models, or closely related areas.
Proven track record of research impact, through publications or applied contributions to production systems.
Hands-on experience building and evaluating large models, particularly those grounded in perception or environment-based reasoning.
Proficiency in Python and experience with modern ML frameworks such as PyTorch or TensorFlow.
Strong understanding of environment-agent interaction, simulation frameworks, or embodied AI.
Ability to translate research ideas into high-quality, practical systems.
Excellent communication skills and experience collaborating with cross-functional teams including engineers, designers, and product managers.

Nice To Haves

Experience integrating AI research into game engines or real-time simulations.
Familiarity with Unity or other 3D engines.
Background in reinforcement learning, planning, or agentic systems.
Experience working with multimodal datasets or synthetic data pipelines.
Passion for empowering creators and improving the AI-assisted authoring experience.

Responsibilities

Drive initiatives to improve spatial reasoning in large models, including understanding scene layouts, object relationships, navigation, and 3D dynamics.
Develop and refine world modeling approaches that allow AI systems to build, maintain, and reason over structured representations of complex game environments.
Advance multimodal LLM architectures that integrate text, vision, and spatial signals to support agentic game authoring and interactive creation tools.
Prototype new AI-assisted workflows, such as natural-language scene editing, spatial instruction following, or agent behavior specification.
Collaborate with engineering and product teams to translate research innovations into production-ready features and tools.
Define evaluation strategies and benchmarks grounded in real game-development scenarios to ensure models are robust, reliable, and performant.
Contribute to the long-term vision for Unity’s agentic AI ecosystem, balancing usability, scalability, and creator empowerment.
Mentor scientists and engineers while fostering a collaborative, innovative, and inclusive technical culture.