Direct Robotics and AI Research
Practitioner research and strategy for the AI era

See the buildout through the people building it.

The AI buildout will shape where capital is committed for decades. Getting the underlying shifts right matters.

We build research from practitioners across the AI and robotics supply chain, adding firsthand observation to company disclosure and public reporting.

They report what they see in their own roles. Our analysts turn it into research investors use to build their own AI era strategy.

Cameras and lidar On board compute Torque motors, magnets Roller screw actuators DETAIL A · HARMONIC DRIVE Agentic AI orchestration World models GPUs and TPUs Data center and power TELEMETRY POLICY
What we cover, from the magnets in the motors to the racks training the models. Contributors report only what they see in their own work.
Who reports
The buildout
Data center operators
Power and grid engineers
Chip fab and packaging engineers
The machines
Robotics deployment managers
Robot component buyers
Magnet and materials buyers
The systems
AI agent and enterprise software operators
Lidar, camera and sensor engineers
Driverless fleet operators

A strategy built on observation, not narrative.

Six things practitioners see in the ordinary course of their work. By the time a change reaches aggregate data, most desks have seen it. Ground to stand on when the narrative moves.

01 How long components take to arrive Lead times move when demand outruns capacity. Buyers see it in their own quotes.
02 Where the industry is standardizing Which component approaches engineers are designing into new programs.
03 What reaches production How often pilots reach sustained operation, as operators see in their own facilities.
04 How concentrated supply is Which components have few qualified suppliers, and how long a second takes to qualify. Pricing power sits there.
05 How build costs are trending Component cost direction across generations, from teardown analysis and observable pricing.
06 How long capacity takes to come online Commissioning, qualification and ramp timelines, from the people doing the work.

Strategy for the AI era starts with what is actually being built.

Public sources reach every desk at the same time. Firsthand observation does not.

Request research access