Knowledge capture
Lore.
Every company has lore.
Most of it retires.
Capture what your best people know — while they work, with the phone in their pocket. Structured, searchable, and owned by your business.
The problem
Expertise walks out the door.
A caretaker retires after 25 years. He knows which valve sticks in winter, which door swells when it rains, why the boiler in building C needs a different startup sequence. None of it is in a manual. It leaves with him.
The same happens with turnover, and it happens when a machine replaces a role: the machine starts from zero, because what the person knew was never captured. The best cleaner, the best mason, the best operator — their edge is exactly the part no document holds.
How it works
Show
A skilled worker records while working. Video and voice, with the phone in their pocket or the bodycam they already wear. No proprietary hardware. The worker starts the recording, the worker stops it. They explain what they do the way they would to a new hire standing next to them — including the corrections nobody writes down.
Structure
Lore turns the raw recording into structured, searchable task knowledge: steps, conditions, exceptions, the why behind each move. It lives in the company’s own Kin — not on our servers, not with a manufacturer. The customer holds the keys.
Transfer
New hires learn from it today — searchable, teachable, tied to the actual site and the actual equipment. And when machines arrive, they inherit it. The same knowledge, in a form that grows toward robot-executable skills as industry standards mature.
What Lore is not
Not surveillance.
The worker starts the recording and the worker stops it. Nothing runs in the background. Lore captures what someone chooses to show — the way they would teach a colleague, not the way a camera watches a shift.
Not a robot-tomorrow promise.
A video with voice is not yet a robot-executable skill — for anyone. Robot-executability grows with industry standards like VLA models and LeRobot. What you get today is knowledge retention that pays for itself with humans. How the standards are maturing →
Where it works
Anywhere the real knowledge sits in hands, not in documents.
The timeline
Today
Knowledge retention.
What your best people know becomes searchable and teachable before it retires. New hires learn from the actual expert, on the actual site — not from a generic manual.
Tomorrow
Your robots start at day one.
Capture what your best people know, today. When the robots come, your business starts at day one — your competitor starts at zero.
Questions
Who owns the recordings?
The business does. Everything Lore captures lands in the company’s own Kin — sovereign cloud or on-premise, keys held by the customer. Not by us, not by a robot manufacturer.
Does it need special hardware?
No. The phone in the worker’s pocket or an existing bodycam is enough. There is nothing proprietary to buy or maintain.
Is this robot training data?
It becomes robot-ready as standards mature — and we are honest about the phasing. A video with voice is not a robot-executable skill today. What you get now is structured knowledge retention for humans: searchable, teachable, tied to your site. As VLA models and formats like LeRobot mature, the same knowledge grows into machine-executable form.
How does Lore relate to Kin?
Kin is the layer: a persistent, owned knowledge entity per person, site, or fleet. Lore is the instrument that fills it. Kin is the family. Lore is what the family passes on.
Last updated: July 2026
One AI.
Every machine.
Kin in every device and robot. Lore keeps what your people know. No setup. No apps. No starting over.