Closed beta · 2026

Signals fromthe physical world.

Phygtl builds the behavioral intelligence layer for systems that need to understand how people move, gather, and coordinate in real spaces. Quietly built. Selectively opened.

For investors, partners, and builders working near physical AI, robotics, spatial systems, campuses, and real-world behavior.

What we reveal

Not another simulation. A living record of behavior where the real world still matters.

01

Real-world coordination.

Structured moments where people choose to meet, move, act, and complete something together.

02

Behavioral context.

Not just presence. The surrounding intent, timing, and social pattern that makes a signal useful.

03

Physical AI substrate.

A dataset direction for models that cannot learn the world from web text alone.

> session.type: coordinated_action
> environment: campus / public_space
> participants: opted_in_group
> behavior: movement + timing + intent
> output: anonymized behavioral signal

> status: closed_beta
> access: selective

The next interface is not a screen. It is how people behave when they are actually there.

Physical AI Spatial behavior Campus network Behavioral data