Never lose a meetingdecision again.
AI notes, decisions, and action items for in-person meetings, hybrid rooms, and video calls. No bot. No data leaving your control.
Meetings / Q3 Board Sync
Q3 Board Sync
Sarah Chen · May 28, 2026 · 42 min
Summary
The board approved the Q3 expansion plan, citing supportive market data. Budget must be finalised by Friday, with revised projections due Thursday and legal looped in on the partnership agreement.
Decisions
- ✓Proceed with the Q3 expansion plan
- ✓Budget finalisation deadline set to Friday
Action items
3 open- EPSend revised Q3 projectionsThuHigh
- JWLoop in legal on the partnershipFriMed
- MRConfirm budget with financeFriMed
Participation
Built for trust from the first call
Meetings are broken. Most tools make it worse.
Meeting bots creep people out.
When an AI bot joins the call, everyone notices — and clients wonder who else is listening. Trust drops before the meeting even starts.
In-person meetings fall through the cracks.
Your best conversations happen in the room. But nobody takes good notes, and most AI tools only work on video calls.
Action items vanish into thin air.
Many AI tools fabricate consensus, misattribute decisions, and hallucinate what was agreed — so nobody follows up with confidence.
Transcription accuracy
Bots in your calls
Data you control
Rivals with white-label
How it works
Three steps. Zero friction.
Capture
Place a phone in the room, or let Lexicanon capture your Teams call automatically via webhook. No bot joins. Nothing to install on attendees’ devices.
Understand
Speaker diarization identifies who said what, then structured notes and action items are extracted — grounded in what was actually said, never hallucinated.
Act
Search across every meeting, share summaries, and push action items to Jira or Asana. Never ask “wait, what did we decide?” again.
Everything your meetings need. Nothing they don’t.
Built for IT admins who need deployment simplicity and leaders who need meeting intelligence that actually holds up.
Works in every room
In-person, hybrid, and video calls — one tool for all of them. Drop a device in the room or capture via Teams webhooks.
No bot, no awkwardness
Captures via device microphone or Teams webhooks. No “Lexicanon Bot has joined the meeting.” Nobody has to wonder who’s listening.
Speaker diarization
Knows who said what, even in a room of ten. Pyannote-powered batch processing reaches 90–95% accuracy.
SSO from day one
Entra ID, SAML, and OIDC on every plan — not locked behind an enterprise tier. Security basics shouldn’t cost extra.
Your brand, your product
White-label with a custom domain, logo, and colours. Launch your own meeting assistant — no rival offers this turnkey.
Your AI stack, your choice
BYOK for transcription (Whisper, Deepgram, Azure) and summarization (OpenAI, Claude, local LLM). No lock-in, ever.
The same calm, focused product — end to end.
Transcript
Know exactly who said what
Speaker diarization separates every voice in the room — even ten people at one table — and labels each line. Search across the whole conversation and jump to any moment.
I think we should move forward with the Q3 expansion. The market data supports it.
Agreed — but we need the budget finalised by Friday. Can finance confirm?
I’ll have the revised projections ready by Thursday EOD.
Analysis
Decisions and actions you can trust
Summaries, decisions, action items, and open questions are grounded in what was actually said. No fabricated consensus — every claim links back to the audio.
- DecisionProceed with the Q3 expansion plan
- ActionEmily → revised projections (Thu)
- Open QWhich region leads the rollout?
Every line traces back to a moment in the recording — click to jump there.
Playback
Listen back, speaker by speaker
The waveform is tinted by who’s speaking, with content-aware chapters. Skim a 40-minute meeting in seconds and replay the part that matters.
Self-hosted
Your meetings. Your data. Your infrastructure.
Deploy Lexicanon on your own servers in a single Docker container. Use local Whisper for transcription and a local LLM for summarization — provably zero-knowledge, your data never leaves your network.
Single Docker container
No complex infrastructure. Pull, configure, run.
Air-gapped mode
Fully offline with local Whisper + a local LLM.
Customer-managed keys
You hold the encryption keys. Revoke access anytime.
GDPR & data sovereignty
Built for EU, healthcare, legal, and government.
$ docker run -d \-e TRANSCRIPTION=local_whisper \-e AI_PROVIDER=local_llm \-e ENCRYPTION=customer_kms \lexicanon/server:latest✓ Local Whisper model loaded (large-v3)✓ Local LLM ready (Mistral-7B)→ Zero data leaves this server. Ever.
White-label
Launch your own meeting assistant. In days, not months.
No turnkey white-label meeting assistant exists in the market — until now. Custom domain, logo, colours, and pricing. Your brand, your product, our engine. Built for MSPs, consultancies, and platform builders.
Custom branding
Logo, colours, and name — fully yours.
Custom domain
meetings.yourbrand.com, end to end.
Multi-tenant
Isolated per customer, by design.
Your pricing
40–60% margin space to resell.
Pricing
Honest pricing. However you want to run it.
Three ways to deploy — fully managed in the EU, your own keys, or fully self-hosted. Every one of them undercuts Otter, Avoma, and tl;dv, often by half.
Hosted · your keys
BYOK
We host it. You bring your own AI keys. Your data path, your AI stack, zero lock-in.
billed annually (€8 monthly)
- Unlimited transcription & summarization
- Bring your keys — Whisper, Deepgram, Azure · OpenAI, Claude, OpenRouter
- You pay providers directly — no usage markup from us
- Hybrid: add our managed services à la carte at cost + 15%
- Everything in Cloud Business, SSO included
Fully managed · EU-hosted
Cloud
We run everything — keys, infrastructure, updates, all in the EU. You just show up to the meeting.
Starter · billed annually (€13 monthly)
- EU-hosted — GDPR & data sovereignty built in
- 20 meeting-hours / user / mo, pooled across your team
- Overage at cost — €0.40/hr, no markup
- Summaries, action items, speaker ID & search
- In-person, hybrid, Teams & Zoom capture
- SSO / SAML included
- Business — €16/user/mo: white-label, audit logs, admin controls, 50 hrs/user & priority support
Your infrastructure
Self-hosted
You run it on your own servers. Pay a flat annual license — nothing leaves your network.
flat license · you provide compute + keys
- Team — €1,500/yr · up to 25 users · all updates · email support
- Business — €4,000/yr · up to 100 users · + white-label, SSO, priority support
- Enterprise — custom · unlimited users, SLA, air-gapped
- Single Docker container — pull, configure, run
- 100% of your data stays on your network
All prices in EUR. Annual billing shown — monthly available at a small premium. Cloud overage is billed at provider cost with no markup. BYOK hybrid services bill at cost + 15%. Self-hosted licenses are flat annual fees; you provide your own compute and API keys.
Questions, answered.
No. Never. For Teams, we receive webhooks via your Entra-registered app — no bot participant appears. For in-person meetings, capture happens on your device. No bot ever joins any call.
No. Lexicanon runs in your browser — there’s nothing for you or your attendees to install. For hybrid calls there’s an optional desktop companion that captures the remote participants as a separate, cleaner channel — still no bot, no virtual device. It sharpens diarization when you want it, but it’s never required.
No. Your meeting data is processed and stored for your organization only. We make this a contractual commitment, not just a promise.
Yes. A single Docker container, fully air-gapped if you want, with local Whisper transcription and a local LLM. Your data never leaves your infrastructure.
Pyannote-powered batch processing identifies individual speakers from a single audio source, reaching 90–95% accuracy even in rooms with many people — well above the 80–85% typical of real-time systems.
Yes — BYOK for everything. Choose your transcription backend (local Whisper, AssemblyAI, Deepgram, Azure Speech) and your summarization provider (OpenAI, Claude, OpenRouter, or local LLM).
Lexicanon is a European company — our managed Cloud runs entirely in the EU, so your meeting data stays under EU jurisdiction by default. Add configurable retention with auto-deletion, data processing agreements, and customer-managed encryption keys on Business+ plans. Want full sovereignty? BYOK keeps the keys yours, and self-hosting keeps every byte on your own network.
90–95% with batch processing using Whisper large-v3. Pick the engine that matches your accuracy, cost, and privacy needs.
Yes — full branding, a custom domain, multi-tenant isolation per customer, and your own pricing. No other meeting assistant offers turnkey white-labelling.
Because security shouldn’t be a luxury tax. Every competitor locks SSO behind pricey enterprise tiers. If your org uses Entra ID or SAML, you shouldn’t pay extra for it.
Nothing breaks. Cloud plans pool hours across your whole team (20/user on Starter, 50/user on Business), and most teams never reach the cap. If you do, overage is billed at our actual provider cost — €0.40/hr, no markup.
No. With BYOK you bring your own transcription and AI keys and pay those providers directly — we never mark up your usage. You pay us a flat €6/user/month to host and run the platform. If you’d rather we handle part of it (say, transcription while you bring your own LLM), those managed services are billed at cost + 15%.
See Lexicanon on your meetings.
Book a 30-minute walkthrough. We’ll show capture in the room, speaker-aware analysis, and how it deploys on your terms — then answer every security question your IT team has.
- Reply within 1 business day
- 30-minute walkthrough
- No bot, runs in your browser