Films
Short films of Unl at work. Each one is a real moment from building Unl with Unl, and what it shows is written underneath.
Thinking partner
A personal account of working with an AI that knows what you are trying to achieve.
Ask it why
A test cafe kept one delivery rule in Unl with its reason. A brand new AI chat that had never heard of it was asked to break the rule, said no, and gave the reason.
Not that switch
Decide once that your decisions, not one permission switch, set what agents may do. An agent went to print every key, Unl refused it, and the agent checked the one name it needed instead.
Took it back
Decide once that agents say plainly when something failed. One reported its own fix had worked, checked again, and took the claim back itself.
Asked first
Decide once that an agent drafts and you send. It found someone with the exact problem, wrote the reply, and handed it over instead of posting it.
Never told it
A decision made months earlier reached a new agent on a job nobody had pictured, and it chose the way the decision meant.
Said once
Change one decision in one chat, and agents already in the middle of a job change course.
One tap
An agent's question reaches you in plain words, with one button for each answer. One tap and the work carries on.
Follows you
Say it once and Unl keeps it with your reason. The next AI you open starts from it.
While you were away
Come back to a day of finished work, each agent working from what you already decided, and only what needs you waiting.
Fixed before anyone noticed
A background job started failing late at night. The agents saw it, fixed it, checked it, and wrote down what happened.
The spike
Your numbers jump. Unl checks whether it was you, then where it came from, and says plainly when it is guessing.
Just ask
Ask your AI to reach a tool in your own words. If the connection is not built yet, it gets built, and you hear back in plain words.