Investors — Pre-Seed Open
The data layer
for physical AI.
Our mission is to democratize physical AI and bring an era of true abundance for all. We’re building the platform that lets anyone who builds machines draw on the same diverse data, models and skills.
investors@aimachinae.comThe Thesis
Three shifts, one opening.
01
Hardware is maturing.
Capable robot bodies are getting cheaper and easier to build. More teams can build a machine than ever before.
02
Data is the new barrier.
What those teams can’t build is the real-world data that makes a machine useful. Today it is collected one robot at a time, inside a few labs.
03
Diversity is the unlock.
Research keeps showing that varied people, places and embodiments generalize better than more of the same. A platform that pools them wins.
The Evidence
Diversity beats volume.
~90%
success in unseen environments with unseen objects, once training data spanned 32 environment–object pairs. Adding environments beat adding demonstrations.
Lin et al. — Data Scaling Laws in Imitation Learning for Robotic Manipulation (2024)
~3×
better performance on skills that existed only in other robots’ data, after pooling 22 robot types from 21 institutions into one training set.
Open X-Embodiment Collaboration — Open X-Embodiment: Robotic Learning Datasets and RT-X Models (2023)
~2×
performance in settings seen only in human video, once a model was pre-trained on enough scenes, tasks and embodiments. Learning from people emerged from diversity.
Kareer et al., Physical Intelligence — Emergence of Human to Robot Transfer in VLA Models (2025)
20,854 h
of first-person human video showed predictable log-linear scaling — and the better the model learned from people, the better a 22-DoF robot hand performed.
NVIDIA — EgoScale: Scaling Dexterous Manipulation with Diverse Egocentric Human Data (2026)
PUBLISHED RESEARCH BY THE GROUPS NAMED. AI MACHINAE IS NOT AFFILIATED WITH THEM.
How It Grows
A platform, not
a robot company.
Every contributor, partner and machine adds to the same shared platform. Value sits in the layer everyone builds on — not in any single machine.
01
Skills marketplace
On-demand skills and utility packages that any machine subscribed to the platform can call.
02
Hardware partners
Machine and capture-hardware makers integrate and certify against the platform.
03
First-party machinae
Our own capture kit and machines — reference hardware that sets the standard others build to.
Why AI MACHINAE
Built for the shape of the problem.
01
Diversity by design
Capture is built around many people, trades, places and devices from the start — the property the research says matters most.
02
Worn, not mounted
Data collection scales with people, not robots. Capture gear that looks like clothing fits into work people already do.
03
Open to every embodiment
Any machine built to the platform draws on the same dataset and skills — so the ecosystem grows without us building every body.
04
We own the platform
The schema, curation, models and marketplace sit with AI MACHINAE. Every new partner and contributor adds to the same platform.
Where We Are
Early, and deliberate about it.
NOW
Capture kit — VISUS, TACTUS, GRESSUS — in design. NEXUM architecture in development. IRIS V1 prototype. FABER HAND in design.
NEXT
First contributors wearing the kit. A shared dataset in one schema. First hardware partners integrating with the platform.
THEN
Foundation models trained on that data. The skills marketplace opens to every machine built to the platform.
White Paper
The full picture, on request.
The NEXUM white paper covers the platform architecture, the capture program, the research behind the thesis, and the platform economics.
Request the white paper
Get in Touch
Interested in investing?
Reach out directly. We respond to all serious inquiries within 48 hours.