Research Engineer, Multimodal Data

About the role

ABOUT EVENTUAL

Robots and world models learn about the physical world from video: millions of hours of people cooking, building, carrying and fixing things. That footage is piling up faster than anyone can review it, and much of it is mislabeled, out of sync or not useful for training. Most teams still choose what to train on by having people watch a small sample. When a model learns from bad examples, it learns the wrong behavior.

Eventual is building data curation for Physical AI. We process video and sensor data at petabyte scale and extract signals from it: how hands and objects interact, what happens over time (a failed grasp, a retry, a recovery) and whether a clip is reliable enough to learn from. Leading Physical AI labs and robotics teams use these signals to decide what goes into their next training run.

The work spans large-scale data systems, computer vision and ML research, and it ships directly into how frontier models get trained. We’re a small team from AWS, Lyft and Tesla, backed by $30M from Felicis, CRV, Y Combinator and the co-founders of Databricks and Perplexity. We helped power the last generation of Physical AI in self-driving, and now we’re building what the next generation trains on.

Join our small (but powerful!) team, 4 days/week in our SF Mission District office.

YOUR ROLE

As a Research Engineer on the Visual Understanding team, you'll own the layer that makes petabytes of video queryable by content. Physical AI teams have video, lidar, radar, and sim outputs scattered across object stores with no way to find what they need without weeks of human annotation. Eventual runs vision/language models and pipelines over every clip in a corpus along axes the customer cares about (gripper type, failure mode, object class, scene, motion density), so a researcher can ask "left-arm grasp failures on deformable objects" and get a curated dataset in minutes.

You'll define the roadmap for our visual understanding capabilities, train and select the models that make corpus-scale annotation tractable at single-digit cents per hour of video, and build the rich datasets that go on to train customer models. This is a applied research role — meaning you'll read papers and run experiments, but you ship to production and your work has a real impact on our customers’ models and robots.

KEY RESPONSIBILITIES

- Own the visual understanding roadmap end-to-end: from picking the model family for a customer's taxonomy to landing it in production inference at corpus scale.

- Train, fine-tune, and evaluate VLMs, VQA models, embedding models, and CNNs against customer datasets and benchmarks.

- Drive down per-clip annotation cost — model selection, distillation, batching, decode pipelining — so "annotate every clip in a 10K-hour corpus" stays economical.

- Build the rich, queryable datasets that customers train on: design taxonomies with researchers, instrument quality, version the outputs.

- Partner with the dataloading and storage teams so visual understanding outputs flow into the index and on to the GPU without re-engineering.