Skip to content
Back to the Journal

AI Implementation

Build Your AI Middle Manager: Tracking, Chasing and Reporting on Autopilot

By Will Ericksson
A feedback loop between people, work and reporting.

In a business of five to thirty people there is rarely a management layer. There is a founder, a few senior people, and a shared hope that everyone remembers what they said they would do. Coordination happens in a Monday meeting and then decays through the week.

This is the gap an AI manager fills. Not leadership. The administrative spine of management that small teams go without.

A week with one in place

WhenWhat the AI manager doesWho acts
Each morningPosts a short status per person: due today, overdue, blockedEach team member
Mid-weekNotices work with no movement and asks the owner, privately, what is in the wayTask owner
As neededSuggests a smaller next action for anything stalledTask owner
ThursdayDrafts nudges for outstanding client or supplier inputsA person approves and sends
FridaySummarises for leadership: delivered, late, at risk, decisions requiredFounder or lead
Month endRuns the recurring process checks: reconciliations done, renewals coming, reports filedProcess owner

We estimate that this kind of tracking, prompting and summarising makes up 50 to 80 percent of what a coordinating manager does all week. It is the pattern we are building into our own group, and the basis of the AI manager engagement.

What it needs

  • A task board or project tool the team already uses, where commitments are recorded by the people who made them
  • Read access to calendars and the shared inbox relevant to the work
  • A channel to speak in: chat, email, or a daily document
  • A clear list of recurring processes with their checks and owners

It does not need a new platform. If commitments currently live in people’s heads, the first job is agreeing where to write them down.

Rules of engagement

We set these with the team present, because acceptance depends on them.

  • Work, not people. It reports on tasks and processes. It does not score, rank or assess individuals, and it produces nothing for performance reviews.
  • Private before public. A stalled item is raised with its owner before it appears in any summary.
  • Humans write the commitments. The system never invents tasks for people.
  • Outbound is approved. Anything to a client, supplier or the wider team is drafted for a person to send.
  • Everything is logged, so anyone can see what it said and why.

What changes for the humans who manage

Senior people stop spending Monday reconstructing the state of the business and Friday writing it up. Their time goes to the conversations a system cannot have: why something keeps slipping, whether the plan is still right, how someone is going.

Starting small

Choose one team and one recurring process. Run the morning status and the Friday summary for two weeks with the team’s feedback. Adjust the tone, because tone is most of what makes it welcome. Then widen.

Tell us what falls through the cracks and we will tell you whether an AI manager would have caught it.

If this sparked something, let's talk.

No pitch, no pressure — just a conversation about what you're working on.

Let's talk
Share: