For IT and admins

We measured the speech-to-text services of the three cloud gateways an InterMIND organization can choose — Azure AI Speech, Google Cloud Speech-to-Text (Chirp 3) and Amazon Transcribe — on the same 153 voice-note clips in 17 languages, one harness, one day. Word error rates per language, the method, the limits of each API, and why the recogniser follows the gateway instead of the leaderboard.

Every InterMIND allowance is metered per meeting host: minutes of voice translation in the meetings you created, words of chat translation for your participants, documents translated, and the team's shared storage. The Billing page shows each meter over a rolling 30-day window, and an email arrives when you approach the storage limit or first reach the translation-minute limit. What each meter counts, what happens at the limit — the meeting continues, translation pauses — and how the 30-day window refills on its own.

Every meeting tool now ships AI, and every AI feature adds a sub-processor. Here is how InterMIND runs its AI features on the hyperscaler gateway your organization already uses — Azure OpenAI in the EU Data Zone, Google Vertex AI in the EU, Amazon Bedrock in your own AWS account — with a per-organization off switch, and what procurement can verify.

You say things in meetings you'd never put in writing — then the meeting runs through clouds, models and retention policies you've never read. Three concrete questions that turn 'confidential' from a homepage word into something you can check, and our answers to each.

A vendor's 'GDPR-compliant' badge means nothing without the work behind it. We ran a full audit of our own codebase against the GDPR obligations that fall on a data processor — erasure, retention, sub-processors, EU runtime — and here's each item we closed, verified against the code.

"GDPR-compliant" on a video tool's homepage means less than you think. Here's the plain-terms checklist of what it actually requires, where Zoom needs configuration, and an EU-runtime alternative built to translate.

A companion to our runtime map: not where your meeting runs, but what it's built from. The layer-by-layer stack — where we run our own code or open-source software, where we're pragmatic about proprietary SaaS, and why the engine most of your data flows through is our own code, with a public, BSD-licensed client SDK.

A vendor-by-vendor map of which external services touch your meeting, where they execute, and what data passes through each — including the AI steps, which run in an EU region of the AI gateway your organization selects.