Growth teams
For a team where more than one person decides
The hard part of a team is not the work, it is the record: who decided, on what basis, and what came of it. Xyranik keeps that record because it is the same record it uses to decide next time.
14-day evaluation · no card · start before you connect anything
What this looks like end to end
Your own systems on the left of each step, the work Xyranik does in the middle, and the outcome it can actually measure at the end.
What changes
How this runs today
The decision lives in a thread
Who agreed, and on what basis, is somewhere in a message. A month later the reasoning has to be reconstructed from memory by whoever is still here.
Things wait for approval in four places
A send sits in one tool, a budget change in another, a post in a third. Something going out that nobody meant to approve is a matter of time rather than luck.
The weekly update is assembled by hand
Somebody spends Friday afternoon pulling figures into a document, and by Monday the numbers in it have moved.
With Xyranik
Four roles, checked on the server
Owner, admin, member and viewer. The check runs where the change happens, not in the browser, so a viewer who finds the URL still cannot change anything.
One place where things wait for a person
Sends that need approval, changes an agent proposed, and decisions that need a human all arrive in one Inbox rather than four. What is waiting is visible; an empty list means nothing is waiting, and is never dressed up as good news.
A weekly briefing everyone gets
One email a week: what happened, what it cost, what changed and what needs somebody. It is claimed once per person per week, so nobody gets two and nobody gets none.
A decision record nobody can rewrite
Every recommendation is stored with what it predicted and the state of the world when it was made. It is scored once, at its horizon. Nothing edits it afterwards, including us.
Autonomy that has to be earned, and can be taken back
Agents start with nothing and are given specific things they may do. After enough scored decisions, an agent that has been getting it wrong has its own authority lowered automatically. It is never raised automatically.
What it does not do for you
Listed here rather than discovered in week two.
No per-person approval routing yet
Anything needing approval goes to whoever has the role, not to a named reviewer or a chain of them.
No organisation-wide two-factor requirement
Each person can turn on a code from an authenticator app, and it is enforced on the server. What does not exist is an admin switch making it compulsory for everyone — and a switch that claimed to enforce something the server did not check would be worse than not having one.
Start before you connect anything
The evaluation begins with what Xyranik can read from your own site, so you can judge what it would propose before deciding whether it may do any of it.