Software spend has an owner. Cloud spend has an owner. The thousands of hours your company spends in meetings every week has nobody — no line item, no review, no way to see waste. Quorum prices every meeting, scores its health, and shows each organizer who actually needs to be in the room.
Metadata-first. Transcript-optional. No bot ever joins your calls.
No agents to install, no behavior change to ask of anyone. Quorum reads what your platforms already record.
Teams/Outlook, Google Calendar/Meet, Zoom — guided in-product, with administrator consent and provider-specific permissions. Setup time depends on your IT environment.
Every meeting deduplicated, priced with your cost bands, scored 0–100 for health. One canonical meeting per real meeting — never 30 copies from 30 calendars.
Organizers review attendance recommendations in Slack. They decide who needs to join and share outcomes with anyone attending asynchronously.
A leadership memo names the most expensive low-health series and what fixing each is worth. Estimated opportunities and observed cost changes are reported separately.
Recording bots see one meeting. Platform APIs see all of them — who joined, when they left, how the same series behaves across months. That's the evidence that survives a CFO's questions.
Join/leave intervals from Teams attendance reports, Meet participant sessions, and Zoom participant records become per-person timelines: who arrived late, left early, or sat through an hour they didn't need.
Flip one switch and speaker-attributed transcripts flow from each platform's own API. Talk time becomes evidence — with a standing rule that a silent approver is still an approver.
A one-tap pulse after each meeting: worth it, or not? Attendee sentiment feeds every recommendation, so the model learns your company, not a benchmark.
Two teams can spend the same hours and have completely different weeks. Quorum measures where the meetings land, not just what they bill — because a day with no unbroken interval in it has a problem that never appears in any total.
The share of days containing an uninterrupted two-hour stretch. Measured on each person’s own calendar and reported per team — never per person.
Distinct meeting blocks, after overlaps are merged. Four scattered half-hours cost what two back-to-back hours cost and leave nothing behind.
Set an invitee cap, a duration cap, an agenda requirement and a protected window, and see how often the calendar keeps them. Quorum measures adherence and never blocks a meeting.
Every invitee classified with a reason attached. Approvers and decision owners are never removed. When evidence is thin, the answer is “unknown” — not “skip it”.
AI recommends — the organizer decides. Both are stored, so recommendation quality is measurable against real outcomes.
Async is not free. A decision that has been dragged across four days and sixty messages has already cost more than the call that would have closed it — and it is still open. Quorum watches for that and says so.
A thread that has run past four hours, involves three to twelve people, has been touched in the last two days, carries no resolution, and has overspent the call. All five, or nothing happens — a governance tool that nags gets muted in a week.
The team’s manager, once, in Slack — what the thread has cost, what the call would, and two buttons. Not everyone in the thread: twelve DMs is more interruption than the thread was causing. Neither button books anything, because Quorum holds read-only calendar scopes.
GitHub webhook analysis stores conversation counts and explicit directory mappings; comment text is not processed or stored. Slack delivers recommendations and surveys. No individual message counts or communication rankings are produced.
Drag the assumptions. Rough numbers are enough — the point is the order of magnitude, and most executives do not finish this calm.
Your scenario assumption; not a measured customer result.
Shortening, splitting or sunsetting standing meetings.
$382,500 a week goes into meetings for 500 people. On these assumptions $61,200 is the modeled weekly cost reduction — $3,182,400 over 52 identical weeks.
This illustration assumes a 40% overlap discount on the recurring reduction and caps the result at total spend. Those assumptions are not validated outcomes. It does not predict cash savings or financial return; actual results depend on implemented changes and data coverage.
Org, team, and series-level analytics only. No individual rankings — enforced in code and in the AI prompts, audited in the repo.
Metadata-first. Sensitive meetings stripped at ingestion. Transcripts off until an admin enables them, purged on a 30-day clock. Retention enforced by a job, not a promise.
AES-256-GCM for every provider token, signed requests on every Slack interaction, company-scoped queries on every route.
The sub-processor list, the control summary and our compliance status are public — no NDA, no sales call first. We state plainly that there is no SOC 2 report yet rather than implying one.
Quorum reads calendar data without editing events. Plan time for administrator consent, an explanation to employees, data verification, and an owner for meeting changes.
A tenant admin runs one provisioning script per platform and pastes the credentials it prints. Read-only scopes. No agent, no browser extension, no bot in any call.
Four or five bands with a loaded hourly rate. This is the only input that turns hours into money, and it stays entirely under your control — nothing is inferred and no payroll system is touched.
Thirty days of history are priced and deduplicated. You see committed spend, attended spend, and the gap between them, by department and by recurring series.
Each recommendation goes to the person who owns the meeting, in Slack, with the money attached. Nothing changes until a human accepts it — which is also what makes the savings measurable afterwards.
Employees are never asked to log in, install anything, or change how they work. If your works council or employee representatives need to review it first, ask us for the briefing pack. Organization, team and series analytics are complemented by scoped person views for workload planning, with no employee performance scores.
Per analyzed calendar user — not per recorder seat. You're buying a cost-reduction system, not another note-taker license.
100–500 users. Agree data access, a baseline and success criteria before starting. Review prioritized opportunities and measured outcomes with leadership. Fees and any annual-plan credit are agreed in writing.
Discuss a pilotMetadata analytics, cost and health, invite optimization, Slack delivery, weekly leadership audit.
Explore a workspaceProposed tier for transcript-backed evidence and additional admin controls. Automatic recaps are not available; scope and availability are agreed during a pilot.
Talk to usPaid plans are agreed by written order and invoice; creating a workspace does not activate a paid subscription. Enterprise (1,000+ users): scoped annual agreements. Storage region, SSO, support and rollout requirements are confirmed during technical validation. Ask for the works-council briefing pack — the document you hand your employee representatives before rollout.
Plan a 4–6 week pilot with read-only access and metadata-first analysis. Agree the baseline, supported integrations, change owners, and outcome review before starting. Scope, fees, and deliverables require a written agreement.
Attended hours times each person's loaded hourly rate, from your own cost bands. We also show a fully-loaded view that adds organizer prep and follow-up time. Every number is deterministic code you can audit — AI never touches the math.
No. Cost, health, attendance patterns, and most recommendations come from calendar and attendance metadata alone. Transcripts are a separate switch your admin controls, fetched through each platform's own API — we never put a bot in your calls.
Three evidence layers. Join/leave intervals from Teams, Meet, and Zoom show exactly who was present, when, and for how long. One-tap pulse surveys capture whether attendees found it worth their time. And if you enable transcripts, speaker-attributed segments add who actually contributed — with a hard rule that silence never marks someone unnecessary.
No. Quorum supports meeting decisions and workload planning. Admins and authorized managers can see calendar commitments and factual participation within their scope. It provides no employee performance score or ranking, and excludes Meet-derived attendance from named person analytics.
No. AI recommends, the organizer decides with one tap in Slack, and we measure what happened. Both the recommendation and the human decision are stored — that audit trail is the product.
They don't. Sensitive meetings are stripped at ingestion — no participants, no attendance, no cost, no AI. Not hidden in the UI. Absent from the database.
Connect a calendar provider in the afternoon; read your baseline the next morning. Cost, top recurring offenders, and async-safe invitees need about 24 hours of sync — no transcripts, no rollout project.