The meeting is not the work
Every chief operating officer (COO) knows the pattern. A meeting reaches a decision and everyone leaves agreeing.
A week later the same people meet again. They spend the first part of it working out what was agreed, who was meant to do it and why it has not happened.
The decision was sound. The follow-up was nobody's job. Someone took notes in their own shorthand. The actions lived in memory. The reminders depended on whoever worried most about the deadline.
That is the thesis: human creativity is the catalyst. A meeting should spend its people on judgement, and hand the work before and after it to AI agents.
Where meetings lose their value
Meetings rarely fail in the room. They fail on either side of it.
Before: the intake. People arrive with different versions of the facts. The first part of the meeting goes on getting everyone to the same page.
During: the minutes. Whoever takes notes is half out of the conversation. Their notes record what was said, not what was decided, by whom and by when.
After: the follow-up. Actions drift. Nobody owns the chase. The next meeting rediscovers them.
None of these steps needs creativity. All three need consistency, which people find dull and AI agents do not.
What AI agents take on
Intake. Before the meeting, an AI agent gathers the current status from the systems where the work happens, the open actions from last time and the questions that need a decision. Everyone receives one short brief.
Minutes. During the meeting, the agent records decisions, owners and dates rather than a transcript. Each action is written so its owner can act on it without asking what it meant.
Follow-up. After the meeting, each action becomes a tracked task with an owner and a date. The agent checks progress, reminds the owner and raises anything slipping before the next meeting. That meeting then starts with what moved, not with what was forgotten.
The person who used to take the minutes is back in the conversation. That alone changes the meeting.
What stays human
The decision itself stays with people: priorities, trade-offs, and anything touching staff, money, partners or legal risk. An agent can remind someone of a commitment. It cannot change one.
The agent's minutes are a draft. The chair confirms them before any action goes out. Everyone in the meeting should know an AI agent is taking the notes and where those notes go.
Autonomy is earned. Start with the agent drafting minutes for the chair to approve. Let it send reminders on its own only once its drafts are consistently right.
Why this is a COO question
Meetings are where an operation coordinates itself. When follow-up depends on goodwill, the whole operation moves at the speed of the busiest inbox.
The value is easy to look for and needs no forecast. Compare a few meetings before and after:
- How many actions from the last meeting were done by the next one?
- How many had to be rediscovered?
- How much meeting time went on status updates rather than decisions?
These are a way to measure, not a promised result. The answers belong to your business.
Questions for your next leadership meeting
- Can we name last meeting's actions, each with an owner and a date, without checking anyone's notebook?
- Who takes the minutes, and what does that cost us in their attention?
- How much of this meeting repeats status our systems already hold?
- When an action slips, who notices, and when?
- Which decisions in this room must always stay with a person?
If the honest answers depend on one diligent person, the meeting only works while that person is in it.
Build, implement and run
Futuvara is a creative technology agency. We build, implement and run AI platforms for our partners: one platform, connected to how your business works, where AI agents handle defined work under your rules and people approve what matters. We implement it inside your live operations and run it so it keeps delivering.
If your meetings make good decisions that never quite happen, that is a conversation worth having.
Is your follow-up nobody's job? Start an Agentic conversation →
