2 Meeting Operations AI Agents, Live in Production.
2 live agents automate meeting operations within operations. They all run on demand. Each one publishes the inputs it needs, the steps it works through and what it hands back.
-
Calendar Invite Generator
Turn a plain-language request into a calendar invite with an agenda, ready for your approval. Nothing is sent.
-
Meeting Notes to Actions
Turn rough meeting notes into decisions, owned action items, and a follow-up invite — separating what was agreed from what was only discussed.
Separating what was agreed from what was only discussed
Meeting notes conflate three things that need to stay apart: what was decided, what someone will do, and what was merely raised. Written up quickly afterwards, they flatten into a single list, and both failure modes follow. Actions appear without owners, so nobody does them and everyone assumes somebody else is. And discussion items get recorded in the same register as decisions, so an idea floated for consideration reappears three weeks later as a commitment the team is apparently behind on — which is corrosive, because it means the notes cannot be trusted and people stop reading them. The alternative, having someone write the notes properly during the meeting, costs the attention of a participant.
These agents handle the two mechanical parts. Meeting Notes to Actions turns rough notes into decisions, owned action items and a follow-up invite, separating what was agreed from what was only discussed. Making the separation structural rather than a matter of the note-taker’s care is what makes the output reliable. Calendar Invite Generator turns a plain-language request into a calendar invite with an agenda, ready for approval — nothing is sent. The agenda is the useful part: an invite without one is the main reason a meeting has no shape, and writing one is the step that gets skipped when the invite is created in a hurry between other meetings.
What this moves
- Actions leaving a meeting with a named owner
- Decisions and owned action items are extracted from the notes, so an action without an owner is visible as a gap rather than assumed to be someone's.
- Commitments nobody actually agreed to
- What was agreed is kept separate from what was only discussed, which prevents an idea raised in passing being recorded as a decision.
Operations
How AI agents handle meeting operations
Drawn from the 2 agents above — what they require, how they run, and what comes back.
What they need
- What's the meeting?
- Invite settings
- Meeting notes
- Follow-up settings
What comes back
- Calendar invite
- What I had to assume
- Agenda
- Recap
- Action items
- Decisions made
- Left unresolved
- Follow-up invite
How they run
- Run on demand
- 2
- Steps per run
- 3
- Credits per run
- 3–4
Where meeting operations fits in operations
Meeting follow-ups, data quality checks and renewal tracking fall between departments, so they slip. Agents are a good fit precisely because the task is well defined and the owner is not.
Next Step
Deploying meeting operations agents
These run as-is against the inputs listed above. Most deployments adapt one — a different source system, a different tolerance, a different approval path. The first call establishes which.