The next ten years of AI will not be won in software. They will be won in the physical world. In factories, hospitals, kitchens, fields, and homes. The companies that own that data will own the century.

microagi is building it. We are the data layer for physical AI.

You will own the depth stage of our annotation pipeline. Monocular and stereo, on footage shot from the head of someone who is moving all day. The hard part is not a convincing depth map on a single frame. It is metric depth that stays consistent across an entire sequence, so the geometry still holds when a model trains on it.

What You Will Do

• Own depth end to end. Model choice, training, evaluation, and the version that runs in production.
• Build monocular and stereo depth that produces metric output, not relative output.
• Hold scale and geometry steady across long sequences, through motion blur, rolling shutter, and changing light.
• Fuse depth with stereo baselines, IMU, and SLAM output wherever it improves the result.
• Build the evaluation sets that catch drift before a customer does.
• Ship improvements into the pipeline and measure what they changed.

Requirements

• Hands-on depth estimation experience, monocular or stereo, at research or production depth.
• Strong multi-view geometry. Calibration, rectification, triangulation, and the failure mode of each.
• Strong Python and PyTorch.
• You have worked with real camera data, not only benchmark datasets.
• Comfortable owning a pipeline stage and being the person accountable for its quality.
• High agency - you don't wait to be told what to do.
• Fluent in English.

Nice to Have

• Experience with metric depth models such as Depth Anything, UniDepth, or Metric3D.
• Temporal or video depth work.
• Rolling-shutter and wide-angle camera models.

#LI-DNI