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 raw footage, lidar, radar, and sim outputs scattered across object stores with no way to find what they need without weeks of human annotation. We change that economics: we run vision-language models 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 research engineering role - meaning you'll read papers and run experiments, but you ship to production and your work is judged by what it does for customer training runs.

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 convolutional perception models 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.
• Work directly with researchers at our partner labs - your shortest feedback loop is their next training iteration.

What we look for

• Strong familiarity with modern vision and multimodal models - convolution nets, VLMs, VQA, embeddings - and a sense for the SOTA that's actually deployable today vs. on a leaderboard.
• Experience running these models at scale on real video and sensor data, ideally for perception tasks (detection, tracking, segmentation, retrieval, captioning).
• Background from a perception team at a self-driving, robotics, or visual-data company - or equivalent depth from a research lab.
• Comfortable with cloud infrastructure and large-scale data processing - you don't need to be a distributed-systems engineer, but you've shipped jobs that ran on thousands of GPU-hours of video.
• Bias toward data and infrastructure: you reach for "annotate the whole corpus" before "fine-tune another model."

Nice to have

• Experience training vision or multimodal models from scratch (not just calling APIs).
• ML/AI research background - papers, citations, or a research org on your resume.
• Hands-on time with big-data frameworks like Spark, Ray, or Daft.
• Worked on embeddings, retrieval, or content-aware search at scale.
• Experience designing labeling taxonomies or running annotation programs.

Perks & Benefits

• In-person, tight-knit team - 4 days/week in our SF Mission office.
• Competitive comp and meaningful startup equity.
• Catered lunches and dinners for SF employees.
• Commuter benefit.
• Team-building events and poker nights.
• Health, vision, and dental coverage.
• Flexible PTO.
• Latest Apple equipment.
• 401(k) plan with match.

If you're excited about being on the team that turns petabytes of raw video into the training data for the next generation of Physical AI, we'd love to talk.