Vue components
A full room, stage, controls, chat, participants, waiting room, captions and AI panel, accessible and themable. Also as custom elements.
EU servers AI with EU providers
Video, audio and chat with an AI that listens, recommends while people talk and, when they hang up, leaves the summary and the proposed actions in your system. With a library and an API key.
Free with test keysNo cardMIT packages
Just want to meet with your team? Use it without integrating anything
Animated illustration of a sales call. Daniel, the customer, says it is more expensive than the competition; the AI privately suggests to Marta, the sales rep, that she answers with the twelve months of support included in the proposal. Daniel says they need to decide before the end of the month and a detector flags it. When they hang up, the AI leaves the summary, the next step, a quote ready to review and a follow-up task.
Illustration with sample data of MeetingRoom, LiveCaptions and AiLivePanel from @iaconsultinglabs/comms-vue.
Integrate in minutes
Your backend creates the meeting with the API key and your screen renders the room with one component. The key never leaves your server.
comms-server.<MeetingRoom>, captions and the AI panel.import { createCommsClient } from '@iaconsultinglabs/comms-server'
import type { Device } from '@iaconsultinglabs/comms-contracts'
const comms = createCommsClient({
baseUrl: 'https://api.comms.iaconsultinglabs.com/v1',
apiKey: process.env.COMMS_API_KEY ?? '', // iacl_test_… to try it, iacl_live_… in production
})
// A meeting with live AI and AI at the end.
const meeting = await comms.createMeeting({
type: 'meeting',
title: 'Customer review',
settings: { guestsAllowed: true, ai: { enabled: true, realtime: true, postMeeting: true } },
})
// Each person's access: your frontend asks for it and you decide.
export function tokenFor(device: Device) {
return comms.issueMemberToken(meeting.id, {
person: { id: 'usr_42', name: 'Ana' },
device,
role: 'host',
})
}
<script setup lang="ts">
import { createMeeting, currentDevice } from '@iaconsultinglabs/comms-core'
import { MeetingRoom } from '@iaconsultinglabs/comms-vue'
import '@iaconsultinglabs/comms-vue/style.css'
const props = defineProps<{ meetingId: string }>()
const meeting = createMeeting({
apiUrl: 'https://api.comms.iaconsultinglabs.com/v1',
tokenSource: async () => {
const response = await fetch(`/api/meetings/${props.meetingId}/token`, {
method: 'POST',
headers: { 'content-type': 'application/json' },
body: JSON.stringify({ device: currentDevice() }),
})
if (!response.ok) throw new Error(`token: ${response.status}`)
return response.json()
},
})
</script>
<template>
<MeetingRoom :meeting="meeting" meeting-title="Customer review" locale="en" captions ai-panel />
</template>
What it is
IACL Comms puts AI meetings inside your product, with your brand and wired to your automations. Your users stay in your app and the results land in your systems.
A full room, stage, controls, chat, participants, waiting room, captions and AI panel, accessible and themable. Also as custom elements.
A REST API with an OpenAPI contract, a Node SDK and signed webhooks. Your backend creates meetings, grants access and receives the results.
GENERAL ERP, our ERP, integrates with Comms as one more customer, with its own key: real proof that it works without shortcuts.
Live AI
Every time someone finishes speaking, the AI processes it together with the context you gave it and everything said so far. If the AI fails or is slow, the call carries on as normal.
Your backend passes the starting context: the customer, the project, quotes, earlier meetings. Text or documents, wherever they come from.
Only people with the AI on see it; guests never do.
“Tell me if the customer mentions a deadline.” “Detect price commitments.” Every match arrives as a typed event, in the SDK and by webhook, to drive your own flows.
You choose what it listens to: the audio, the meeting's public chat or both. Private messages are never processed, and anyone who asks is left out of the AI.
AI at the end
Who spoke and when, segment by segment, delivered to your system.
With the sections you define in each mode: minutes, decisions, commitments, next steps…
Tasks, quotes or follow-ups of the types your system knows how to run, with the sentence that backs them. Comms proposes; your system approves and runs them.
Each company defines its modes (sales, site follow-up, interview…) with their checks and their results as validated JSON.
Use cases
For developers
Your backend creates the meeting with the API key and gives each person their access. The browser joins with that access: the key never leaves your server.
@iaconsultinglabs/comms-server for your backend and @iaconsultinglabs/comms-vue for the screen, under the MIT licence.iacl_test_…To build at no cost: meetings of up to 4 people and 15 minutes, and a simulated AI that answers the same every time, without calling any provider. Their configuration and usage are kept apart from production.
iacl_live_…With the AI models of the catalogue and their price. Each key with its permissions and limits, rotation without downtime and usage per key. Shown only once, and we only keep their hash.
Pricing
Minutes are person-minutes: 3 people for 10 minutes are 30. Prices exclude VAT.
Provisional prices. They may change before launch.
To integrate and test at no cost.
€0
Your first integration in production.
€49 / month
More volume and more detectors.
€199 / month
For large volumes or your own infrastructure.
Custom
One plan for everything: with Starter or Pro your team also gets Meetings in the portal (more people, longer meetings, recording and AI), and its usage counts against the same minutes. See the plans for teams.
| Included per month | Sandbox | Starter | Pro |
|---|---|---|---|
| Price | €0 | €49 | €199 |
| Live keys | No | Yes | Yes |
| AI models | Simulated | Catalogue | Catalogue |
| Meeting minutes | 600 | 10,000 | 50,000 |
| Recording minutes | — | 500 | 3,000 |
| Live AI minutes | 300 simulated | 500 | 2,000 |
| AI closings | 50 simulated | 10 | 30 |
| Active detectors | 3 | 5 | 25 |
| Usage beyond the plan | No: stops at the cap | Yes, per minute | Yes, per minute |
| What | Price |
|---|---|
| Meeting person-minute | €0.002 |
| Recording minute | €0.006 |
| Live AI person-minute | €0.03 |
| AI closing | €1 |
Each minute one person spends connected to a meeting. A call of 2 people for 30 minutes is 60 person-minutes.
On Starter and Pro the meeting carries on and the extra is charged at the table price with the monthly invoice. Sandbox has no extra: at the cap, new meetings cannot start until the next month.
No. VAT is added on the invoice according to your country and tax details.
Yes, from the portal and with no commitment. The plan stays active until the end of the period already paid.
Yes. The Sandbox plan gives you test keys and a simulated AI to build everything at no cost and with no card. When you want to go live, you choose a plan.
We let you know and you have 7 days of grace. If it is not sorted, new production meetings are paused; a meeting in progress is never cut.
Create your account, generate a test key in the portal and follow the quickstart.