IO Coach™, say how the practice should run.
Describe the rhythm you want, a playbook, a population, a cadence, an owner, and IO resolves it against the data it already holds, schedules it, generates the work, and measures whether it is actually being adhered to. It is not a chatbot, and it does not decide anything on its own.
It proposes. A person approves. Only then is anything created.
Every proposal derives from data already held in the tenant, populations, playbooks, owners, prior execution. Nothing is scheduled, no work is generated and no owner is assigned until a named person approves it.
Declare the practice
One sentence: position role-based training every two weeks for the roles carrying the most risk; increase supplier reassessment from monthly to fortnightly; run access recertification quarterly for anyone holding standing privilege.
IO resolves it
The statement is resolved to an existing playbook, a real population with a count, a cadence, a next due date and an owning seat, not to a generic reminder.
A person approves
The resolved programme is shown before it exists. Approve it and IO schedules it and generates the work; decline and nothing is created.
Adherence, measured honestly.
A running programme names the population it applies to and how many people are in it, so a cadence is not an aspiration.
A rate is not shown until three consecutive periods have resolved. Before that the programme says so, rather than publishing a figure that cannot yet be trusted.
Accountability is a person and a date, the same rule the rest of the platform runs on.
IO runs with or without Coach.
Coach is an engine you can turn on, not a dependency you are forced into. Every playbook, schedule, population, owner, and verification result in IO exists and runs independently of the AI layer.
If you do not want to use Coach, or cannot use an AI engine for policy or regulatory reasons, the rest of the platform keeps working exactly the same way. You can still declare practices manually, assign owners, run work, and measure adherence.
The second thing it does.
Coach also writes the briefing behind a correlated case, from evidence already held, what was seen, across which systems, and what it appears to add up to. It is drafting on top of existing evidence, not new analysis and not a new conclusion.
It schedules playbooks that already exist. It never decides what a measurement says, verification remains deterministic, and the result stands whether it is good or bad.
See it before you talk to anyone.
Two quarters of recorded activity across sixteen seats. One click, no install.