Verne Robotics
Full-stack physical-AI automation vendor: builds its own bimanual robot arms and grippers, trains in-house manipulation foundation models from teleoperation/wearable/field data, and operates the deployed fleet with continual learning, sold as pay-by-the-hour robot labor rather than capex hardware. · United States
Bimanual robot arms with in-house manipulation foundation models that learn new tasks in hours, deployed as pay-by-the-hour physical labor on U.S. factory floors and in labs.
Nemo3 bimanual robot arms plus a vertically integrated autonomy stack (teleoperation and wearable/egocentric data collection, skill decomposition, diffusion-policy training, on-site adaptation, continual learning from field interventions). Sold as a drop-in, retrainable automation service priced per hour instead of a large capex purchase; deployment in days (4 days at ABClonal) with customer staff able to retrain tasks in hours.
Research mapping
Physical AI / Autonomy Software, Fleet Platforms & End Markets (L10) / Full-stack physical-AI automation vendor: builds its own bimanual robot arms and grippers, trains in-house manipulation foundation models from teleoperation/wearable/field data, and operates the deployed fleet with continual learning, sold as pay-by-the-hour robot labor rather than capex hardware.
Evidence boundary
Original analysis and independent corroboration remain under development. Commercial scale and durable technical advantage remain under review.
What makes it defensible
The deployment gap, not the robot: conventional arms need months of custom engineering and are hard-coded to one workflow, so high-mix back-of-house work stays manual (company cites ~1 in 3 U.S. SMB back-of-house roles unfilled). Verne claims its models turn ~30 minutes of teleoperation data into reusable skills, collapsing integration from weeks to hours and making pay-per-hour robot labor economic at low volume.
Position in the stack: Sits at the autonomy/end-market layer of physical AI and integrates downward: it is simultaneously the arm/gripper integrator, the model trainer and the fleet operator serving end customers, which lets it capture deployment economics and own the field-data flywheel.
Who buys it
Named customers: ABClonal (Boston life-science reagents; named case study, vial/reagent packaging)
Founding team
Source
Financing detail
- Last round
- $500K Pre-Seed (Aug 2025)
Investors: Y Combinator (lead, S25), Crucible Capital, 468 Capital, Greylock, Scale Asia Ventures, J20 Ventures
Mapped funds and investors
Direct relationships supported by published evidence.
Reported holdings as-of
No sourced relationships mapped yet.
Disclosed investment history
No sourced relationships mapped yet.
Showing up to four relationships in each group. ETF weights describe the fund’s portfolio, not ownership of this company. Private investment history does not establish current ownership.