Performance Engineer (Inference, Training & GPU)

About the role

ABOUT WORLD LABS

World Labs https://www.worldlabs.ai/ is a frontier AI research and product company advancing spatial intelligence, the next frontier beyond large language models. Co-founded by Dr. Fei-Fei Li https://profiles.stanford.edu/fei-fei-li, Justin Johnson https://www.linkedin.com/in/justin-johnson-41b43664/ and Ben Mildenhall https://www.linkedin.com/in/ben-mildenhall-86b4739b/, the company is pioneering world models that perceive, generate, reason, and interact with virtual and physical worlds.

The company’s flagship product, Marble https://marble.worldlabs.ai/, transforms text, images, and video into fully navigable 3D worlds, unlocking applications across gaming, film, architecture, robotics, and immersive digital experiences. Backed by leading investors and with over $1B raised, World Labs is assembling a world-class team at the intersection of AI research and real-world deployment.

ROLE OVERVIEW

We are looking for a Performance Engineer to make World Labs’ models train and serve as fast as the hardware allows.

Running large generative world models at scale is a novel systems problem. You will find the bottlenecks — in kernels, in the serving path, in the training loop, in how we use our GPUs — and eliminate them. Your ownership is technical and concrete: the throughput you unlock, the latency you cut, the utilization you win back, and the correctness you hold while doing it. You will work up and down the stack, from low-level tensor and kernel optimization to fleet-wide serving efficiency, in close partnership with the researchers whose models you are accelerating.

This is a hands-on, individual-contributor role. You will profile, design, build, and ship code directly.

WHAT YOU WILL DO

- Optimize inference and serving end to end — latency, throughput, batching, caching, and scheduling — to serve our models efficiently at production scale.

- Write and tune GPU kernels (CUDA, Triton) for hot paths; drive kernel fusion, memory- and bandwidth-bound optimization, and low-precision (FP8/INT8) execution.

- Optimize training throughput and GPU utilization: parallelism strategies, communication/compute overlap, mixed precision, and eliminating pipeline stalls.

- Build performance models, profiling workflows, and observability that make throughput, latency, cost, utilization, and their tradeoffs legible across the stack.

- Own numerical correctness across precision, kernel, and hardware changes — treating correctness as part of performance, not separate from it.

- Partner with researchers to productionize models for serving and to make experiments run faster and more reliably.

- Where needed, work on the distributed systems that training and inference run on — but the core of the job is squeezing the most out of every GPU.

KEY QUALIFICATIONS

You should excel at the fundamentals below — we index on inference, serving, GPU optimization, and training performance. Distributed-systems breadth is welcome, but secondary.