Aktoria Robotics
Robot Learning Researcher (VLA / Foundation Models)
San Francisco, CA, US · Not specified
- Annual base salary
- $250k – $500k USD
- Equity
- Not disclosed
- Commitment
- Full Time
- Company stage
- Not disclosed
Compensation as listed
$250K - $500K • 1.00% - 2.00%
About Aktoria Robotics
Aktoria builds remote teleoperation software for robots. A remote operator can run a robot anywhere in the world as if standing right next to it. Our product is what makes deployed robots useful today.
About the role
-
Train / fine-tune VLAs and robot foundation models
-
Run data-quality ablations and scaling studies
-
Publish papers and benchmarks
-
data curation, filtering, formats, evals on real robots
expected to have:
-
experience training IL/VLA policies at scale
-
First-author publications (CoRL/ICRA/IROS/RSS/NeurIPS) or equivalent strong open-source work
Technology
Remote teleoperation of robots over the public internet. Core problems:
- Remote operations over unreliable networks.
- Low-latency, high-fidelity data streaming for high telepresence.
- A custom transport and networking stack built for real-time control.
- Supporting different robots and hardware in the field, over varying connections.
Source: Y Combinator. Confirm availability with the employer.
Apply through the original posting.
View listing