We’re K-Scale Labs (https://kscale.dev). We’re a team of engineers and researchers building humanoids and a robotics general purpose foundation model in Palo Alto, USA.
In the last six months, we've trained state-of-the-art machine-learning models on our training infrastructure, developed the operating system, designed and manufactured 2 humanoid robots.
In The Next Six Months, We Will Be Pushing The Boundary On Whole-body And Manipulation Models And Manufacturing Our Hardware At Scale.\ \ What You Will Work On
- Whole-body RL policies and manipulation policies.
- Vision-Language Action (VLA) policies
- RL whole-body control policies and sim2real
- K-Sim RL training library: https://github.com/kscalelabs/ksim and AMP
