Meetings, by Unseen Dynamics
Not another notetaker. A validated read of whether your meetings are generative — and why.
Today's AI meeting tools give you a transcript and a summary. Meetings adds the part that changes anything: a structural diagnostic — generative, sterile, or damaging — grounded in a relational model that has been empirically validated, with concrete guidance to move a meeting toward generative.
Early access. The diagnostic is in calibration on real meetings — we are onboarding deliberately, not charging yet.
The difference
Notes are a commodity. A validated diagnostic is not.
Otter, Fireflies, Fathom, Granola — the notes market is crowded and converging. Our differentiator isn't another summary. It's that the read is a productised version of an error-barred research instrument: the same structural detectors that were validated in our science programme, applied to your meeting. It measures structure, not sentiment.
How it works
Connect once
Link your calendar. Choose when the diagnostic runs — every meeting, only the ones you host, only those with external guests, only the ones you tag, or off. Per-meeting on/off any time.
It reads the structure
After a meeting, the engine reads the interaction the way our validated research instrument does — per-participant fidelity, whether the exchange served its stated purpose, and whether the room fell into an unproductive loop.
You get the why
A structural read-out: generative, sterile, or damaging — and the pattern that drove it — plus two or three concrete, structural next steps. Never a verdict on a person.
What it tells you
Generative / sterile / damaging
An overall structural verdict for the meeting, grounded in a research-validated instrument — not a sentiment score.
The loop, named
When two people lock into an unproductive back-and-forth, the diagnostic detects it structurally (the FM2 mutual-recursive-abstraction pattern) — so you can name it and break it.
Who the meeting served
The beneficiary-direction read — the empirically decisive signal — surfaces when a discussion quietly drifts to serve one party rather than the shared goal.
Or: not enough signal
If the meeting doesn't carry a clear structural read, the diagnostic says so. It does not manufacture a confident verdict.
Privacy by construction
Structural, never personal. Yours, never your manager's.
The report describes the interaction's structure— never a verdict on a person's character. Any individual private report is owner-scoped by construction: only you can see your own. Team tiers show meeting-level reads and anonymised org-health aggregates — never surveillance of individuals. Recording and analysing a meeting requires participant consent; the diagnostic runs EU-hosted.
Pricing
Priced on value, not minutes.
Free trial
14 days, no card
Enough real meetings to see the value — several structural reports, full features.
Pro
~€24 / user / mo
For individuals. Unlimited personal meeting diagnostics (fair-use), full report history and structural recommendations.
Team
~€49 / user / mo
For workspaces. Calendar auto-capture, meeting-level team reports, and anonymised org-health aggregates — never surveillance of individuals.
Provisional pricing, to be confirmed before launch. Prices are VAT-exclusive; VAT is added at checkout where applicable. Annual billing ~20% lower than monthly.
See your meetings, structurally.
Join early access while we calibrate on real meetings.
Request early access