SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Trace Labs is building the data infrastructure for physical AI—enabling robotics and embodied systems to see, move, and interact with their environment. The company addresses a critical gap: there's no scalable way to collect high-quality, real-world training data for robotics at the scale that exists for language models. Trace captures how humans interact with the physical world and builds ML systems to maximize the value of that data.
This role sits between ML engineering and research. You'll own machine learning models end-to-end—from data collection and training through evaluation and deployment into the annotation pipeline. You'll work closely with the computer vision team but with a broader, product-focused scope. You won't specialize in a single data type; one month you might work with sensor signals, the next with hand-tracking video, then language data. The emphasis is on getting high-quality models working quickly and shipping them into production.
Key responsibilities include:
- Owning models end-to-end: data, training, evaluation, and deployment
- Building training pipelines for large, multimodal datasets (video, sensor streams, language)
- Partnering with the Head of Engineering and CV team to productionize models and drive continuous improvement
- Starting with the simplest working approach, then iterating; pulling from recent research when it helps you ship
- Moving quickly between different problems and building tooling as needed
The team is early-stage and deeply technical. They value what you've built and shipped over where you've published. If you're earlier in your career but have a clear track record of moving fast and owning hard problems, they want to talk.
Why join: You'll work on a foundational problem—the data layer for physical AI is still being defined. You'll have access to data at scale from day one. Your models ship into production and directly shape what the data can do. There's room to grow deeper into ML, take on bigger systems, or both.
REQUIREMENTS:
- BS in Computer Science, Electrical Engineering, Mathematics, Aerospace, or related field; or equivalent practical experience (MS is a plus, not required)
- 2+ years of hands-on industry experience training deep learning models on real-world data (at a company, startup, or national lab); earlier career acceptable if track record demonstrates it
- Strong proficiency in Python and a modern deep learning framework (PyTorch or JAX)
- Solid grounding in ML fundamentals: architectures, optimization, loss design, evaluation
- Comfort with messy, real-world data (video, time series, sensor streams)
- Extraordinary bias toward shipping—move from idea to working model fast, make good calls under uncertainty
- High agency—find the problem, own it, get it done without waiting to be told
- Range—excited to work across different data types and problems, not just one narrow specialty
