Meeting Intelligence is the part of NearSync that remembers meetings. Not that they happened, which the calendar already knows, but what was said, what was decided, and what was promised.
The parts
Pippin Scribe is the recorder. It joins a call as a visible participant, and everybody can see it there.
Summaries is what you get afterwards: the write-up, the decisions, the action items, and the measured outcomes.
Intelligence is the analysis of the conversation itself: who spoke, how much, who asked the questions, what kept coming back.
Coach, which is a separate tab, is about you rather than about the meeting.
Why it is not just a transcript
A transcript is a wall of text nobody reads twice. Four things happen to it here, and each one solves a different problem.
The summary. What happened, in a few paragraphs, for the person who was not there.
The decisions. What was actually settled, pulled out separately, because a decision buried in the fourth paragraph is a decision that will be relitigated in a fortnight by somebody who genuinely does not remember it.
The action items. What people committed to, with the owner and the deadline where those were said out loud. These become real tasks in one click.
The outcomes. The unusual part, and the reason this is worth configuring properly: the meeting is scored against what that kind of meeting is supposed to achieve.
Alongside those it records the tone of the conversation and the topics that dominated it, which are most useful in aggregate rather than one at a time.
Outcomes, briefly
A discovery call is supposed to establish certain things. Whether there is a budget. Whether you are talking to somebody who can decide. Whether there is a real timeline or a polite one.
Your organisation defines those questions once per meeting type. Every recorded meeting of that type is then measured against them, and the answers sit on the summary as a scorecard.
Over a quarter that turns into something no transcript can give you: which discovery calls actually qualified, run by whom, and what the ones that did had in common. The full mechanics are in the piece on meeting types.
Two firm boundaries
Everyone can see the recorder. It joins as a named participant in the list. There is no hidden recording mode and no configuration that produces one.
Internal measurements stay internal. Outcome scores, coaching, and any analysis of how the conversation went are never included in anything sent to a participant. The write-up a customer receives and the scorecard your team sees are deliberately different documents, and the separation is enforced in the pipeline rather than left to whoever presses send.
That second boundary is what makes honest measurement possible at all. If a score could leak to the customer it was about, nobody would score honestly, and the whole exercise would become theatre within a month.
What it costs
Recording draws on your organisation's included meeting hours, not on your credits. Fifty hours a month on Standard and a hundred on Pro, pooled across everybody rather than allocated per person.
The analysis afterwards costs credits, because it is a language model reading a transcript. A one-hour meeting costs meaningfully more than a chat question and dramatically less than an hour of anybody's time.
Coaching is a separate, on-demand cost, which is exactly why it sits behind a button rather than happening automatically to every meeting.
The Intelligence tab is free, because it is a parse rather than a model. That is also why it works on every meeting you have ever recorded, including the ones from before any of this was set up.
Where to start
Record one real meeting. Read the summary. Then open the Intelligence tab for the same meeting.
Most of the value here is not obvious in the abstract, and the argument for configuring meeting types properly is much easier to make once you have seen the analysis applied to a conversation you actually remember having.
How the pieces fit together
The order matters, and it is worth knowing which decision affects what.
Before the meeting, somebody sets a meeting type on the event. That single choice decides which framework the analysis uses and which outcomes get measured, and it is the only preparation the system needs.
During, the recorder captures. Nothing else happens.
After, the transcript is produced, then analysed. The summary, decisions, action items and outcome scores all come out of that one pass, which is why they arrive together rather than trickling in.
Separately and for free, the transcript is parsed for conversation measurements. That parse is deterministic, so it is available on old meetings too.
On request, coaching is generated for a specific meeting.
The practical consequence: the only thing you can get wrong in advance is the meeting type, and the only thing you have to remember afterwards is to promote the action items.
Who uses which part
Everybody reads summaries. That is the part with universal value and no setup cost.
Anybody running customer conversations should care about outcomes, because that is where the scorecard turns a pile of write-ups into an answerable question about qualification.
Anybody who wants to get better at meetings generates coaching, occasionally and deliberately.
Administrators configure the meeting types, once, and then largely leave it alone.
Nobody has to use all of it. A team that only ever reads summaries and promotes action items is getting most of the value.
What it will not do
It will not check what was said. The write-up faithfully records claims, including the ones that were wrong. It does not compare a quoted figure against your data.
It will not attend for you. A summary is a poor substitute for having been in a conversation where a decision was made about your work.
It will not fix a bad meeting. A conversation with no agenda produces an accurate summary of a conversation with no agenda.
Did this answer your question?
No, ask a person