JARM.AI
jarm builds the humanoid base layer
— hardware, core stack, and developer tooling —
so developers can create tasks, workflows, and real-world applications faster.
share your interest
and get updates ->
just  another  robot  manufacturer
JARM.AI
philosophy
Most humanoid companies optimize for one of two extremes. A single narrow use case, or one large model meant to do everything. Both approaches create friction for developers and make scaling harder.
jarm takes a different path. We build below the use-case layer: the humanoid base, core stack, and developer tooling that make task creation easier. Our focus is not to compete at the application layer, but to enable developers, labs, startups, and frontier teams to build on top of a practical humanoid platform.
The real question is no longer whether humanoids can do useful work. The question is whether they will be usable, adaptable, and scalable enough to matter in real workflows.
developer UX
jarm is built for task creation, not task ownership. We help teams move from data and simulation to reusable tasks, then extend those tasks with components and runtime modules for deployment.
A typical workflow looks like this: Data and simulation → task creation → components and extensions → humanoid deployment
Tasks are modular building blocks inside larger workflows. Examples include picking, grabbing, using tools, and interacting with an environment. This makes it easier to combine capabilities, adapt them to different contexts, and deploy them at runtime.
why jarm
Founder-led and product-focused. jarm is being built by a small team, led by Jacob Munke with a strong focus on usability, adaptability, and humanoid enablement.
Self-funded through prototype 1. That gives the company execution discipline and lets us move with focus before raising externally.
Built in Europe, close to robotics and AI talent. With proximity to THWS and JMU, jarm has access to engineering talent, academic collaboration, and a strong base in Germany and central Europe.