Software Engineer, Robotics Data

San Francisco$130,000 - $500,000

About Mercor

Mercor's mission is to organize human intelligence to power the AI economy. We're a leading AI data company, building the layer between human expertise and frontier models. Millions of domain experts on the platform are paid over $4 million per day to train frontier AI models. Mercor's APEX benchmark family measures AI's real-world impact on professional work. Mercor Enterprise brings this same infrastructure to Fortune 500 companies: helping companies capture how their best people actually work, translating that expertise directly back into agents.

Mercor is creating a new category of work where expertise powers AI advancement. Achieving this requires an ambitious, fast-paced and deeply committed team. You’ll work alongside researchers, operators, and AI companies at the forefront of shaping the systems that are redefining society. Mercor is a profitable Series C company valued at $10 billion. We work in-person five days a week in our San Francisco, NYC, or London offices.

About the Role

Frontier AI is going physical, and the labs building it are bottlenecked on one thing: high-quality data from the real world. Mercor pairs its operational scale with the specialized engineering that physical-world data demands.

You'll build the data backbone of physical AI: the pipelines that take raw multi-sensor capture from the field (video, depth, inertial, audio, and more) and turn it into validated, privacy-safe, delivery-ready datasets for frontier labs. Ingest, segmentation, pre-labeling, automated QC, and packaging, at petabyte scale across thousands of concurrent collectors.

What You'll Do

Build the end-to-end sensor data pipeline: ingest from capture devices in the field, through segmentation, pre-labeling, QC, and packaged delivery to customers
Design automated QC that validates recordings at scale: timing and sync integrity, calibration health, sensor continuity, coverage against requirements
Establish dataset schemas, versioning, provenance, and versioning, so every delivery has a clear system traceability
Build shared processing components such as privacy redaction, transcription, encoding, format packaging across all offerings
Integrate VLM-assisted pre-labeling and quality scoring into production workflows without sacrificing debuggability or human oversight

What Makes This Role Different

High ownership, early. This is a young, strategically central product area; the product you build will shape Mercor’s physical-world data collection standards
The data is the deliverable. The end product at Mercor is the data; what your pipeline produces is what shapes the models that large frontier lab trains on
Real physical-world scale. Your inputs come from devices operated by humans in global real world settings, for thousands of hours. Building systems that scale is precedent.

What We're Looking For

Strong production backend/data engineering experience — you've built and owned high-volume data pipelines
Experience processing video or sensor data at scale: large binary formats, streaming ingestion, distributed batch processing, object storage economics
Fluency in Python and comfortable with AWS
Genuine data taste: you can look at a sensor trace or a timing histogram and tell when something is off
Comfort in ambiguous, fast-moving problem spaces where requirements evolve with the customer

Nice to Have

Experience with robotics data formats and tooling (MCAP, ROS bags, protobuf, Foxglove), camera geometry, or multi-sensor calibration and synchronization
Computer vision or multimodal ML experience (detection, tracking, VLM-based labeling or QC)
Prior work on data engines for AV, robotics, or egocentric video