Eka Robotics

Summary

About

About Eka Robotics

Eka Robotics develops robot-learning systems for dexterous manipulation, combining a Vision-Force-Action model with force-sensitive robotic hands. The Cambridge, Massachusetts company designs its model, sensing, control and hardware together so that contact provides information during a grasp or insertion. Its approach addresses variations such as slipping objects, changing friction and imperfect alignment, with visual perception guiding the approach and force feedback supporting physical interaction.

Eka trains manipulation policies through reinforcement learning in physics simulation, allowing virtual robots to practise variations and recovery behaviours before policies reach hardware. Its demonstrations include fragile-object handling, bin picking and contact-sensitive insertion tasks. It is developing customer deployments, with simulation-trained policies and force-aware hardware forming an integrated manipulation system.

PRICE
Range 1Y
Return
High
Low
Updated

Funding

Investors

  • Andreessen Horowitz - Lead - Angel / Individual - 1 funding event - first seen 2026-10-20
  • Bison Ventures - Participant - VC - 1 funding event - first seen 2026-10-20
  • E14 Fund - Participant - VC - 1 funding event - first seen 2026-10-20
  • F-Prime - Lead - Unknown - 1 funding event - first seen 2026-10-20
  • Glasswing Ventures - Participant - VC - 1 funding event - first seen 2026-10-20
  • MFV Partners - Participant - VC - 1 funding event - first seen 2026-10-20
  • Obvious Ventures - Participant - VC - 1 funding event - first seen 2026-10-20
  • Samsung Next - Participant - Corporate / CVC - 1 funding event - strategic/corporate - first seen 2026-10-20

Team

  • Pulkit Agrawal - Co-founder
  • Tuomas Haarnoja - Co-founder

Timeline

Events

Context

Comparable profiles

Context tags

  • Manipulation foundation model - Capability - Vision-Force-Action model for general-purpose contact-rich manipulation; described demonstrations do not establish sustained production deployment.
  • Reinforcement Learning Control - Stack Layer - Manipulation policies trained with reinforcement learning at scale in physics simulation and transferred to physical robots.