Machenta

Overview

Machenta is a frontier AI company developing foundation models for robotics designed to enable instant, human-like learning [0, 2]. By utilizing a proprietary "vision-demo-action" (VDA) model trained exclusively in simulation, the platform allows robots to learn new tasks from a single demonstration without the need for extensive post-training or real-world data collection [0, 2]. Unlike traditional vision-language-action models that often require days or weeks of data gathering to master specific tasks, Machenta’s system focuses on meta-learning, teaching the robot how to learn rather than memorizing individual motions [2]. Founded by researchers Edward Johns and Vitalis Vosylius—whose academic background in robot learning stems from Imperial College London—the company is building a team of specialists to scale its simulation-based approach [1]. The startup, which is based in London, recently announced a $40 million Seed round to support its development of physical AI [0, 1].

Funding

Total disclosed: $40M across 1 round. Computed only from sourced rounds.

Machenta funding rounds

Date	Company	Amount	Stage	Lead	Other investors	Sector	Sources
Oct 27	Machenta	$40M	Seed	Creandum, Union Square Ventures	—	Robotics & Automation	1

• Oct 27 · Seed

  Machenta$40M

  Led by Creandum, Union Square Ventures

  Robotics & Automation1 source

Investors

• Creandum
• Union Square Ventures