Why we built Scribr on your phone, not your calendar

The meeting-bot problem nobody talks about

When we started building Scribr, the obvious path was to mirror what Descript and Otter had done: calendar integration, meeting bots, enterprise rollout. Otter had already won the market on that front. Descript had the video editor that made transcription feel like design. Both had huge teams and funding.

But we kept hearing the same refrain from users: 'I don't want to invite a bot to my Zoom. I don't want to ask permission from five people in the room. I just want to hit record on my phone.'

That friction is invisible until you feel it. A therapist can't ask a client to approve a meeting bot. A freelancer jumping between calls with three different clients doesn't want to set up integrations for each one. A researcher conducting interviews can't have a robot sitting in the conversation. The moment we understood that, we stopped trying to own the calendar and started owning the pocket.

Mobile-first means you capture first, ask permission later

Descript is built around the laptop. You upload, you edit, you share. It's a beautiful post-production workflow. But it assumes your meeting is already over, already recorded, already on a file somewhere. Scribr assumes you're in the moment, phone in hand, needing to hit a button and start listening instead of typing.

The Quick Record widget changed everything for us during testing. Users could swipe up on their home screen and start recording without unlocking their phone. The Siri shortcut let someone tap their AirPods and say 'Hey Siri, start recording with Scribr' mid-call. Nobody was asking for a feature roadmap. They were asking for speed. Milliseconds matter when you're trying to capture something before it's gone.

On the Free tier, we keep everything local. Whisper and Apple Speech mean your audio never leaves your device. That was non-negotiable. The moment we asked for permission to send voice data to the cloud, we'd lose half our user base on trust alone. So we made the private version fast enough that it doesn't feel like a compromise.

Where cloud transcription actually earns its place

Here's what we won't claim: on-device transcription handles every scenario. If you're transcribing an hour-long interview, or you need action items automatically extracted, or you're trying to find a moment from six months ago, you need real infrastructure. That's where Deepgram comes in on the Pro tier and above.

The difference is why you'd use it. You're not subscribing because transcription is cool. You're subscribing because you have 500 meetings a month and you need AI to find the buried decisions inside them. You're paying for summaries and action items because nobody's got time to re-read 45 minutes of notes looking for 'who's responsible for the budget review.'

Descript makes editing video and audio beautiful. We're making finding what matters in your conversations frictionless. That's a different product for a different moment in your workflow.

Privacy and compliance aren't features. They're promises.

A month after our first legal customer signed up, we learned that our initial compliance story was too loose. They needed audit logs. They needed GDPR modes where certain data never touched cloud storage at all. They needed to know exactly what Scribr could and couldn't see.

We built Vault Mode. AES-GCM encryption. Biometric lock app-wide. The Team tier includes GDPR compliance modes with audit logs because the moment you're handling client data or patient conversations, hand-wavy reassurances aren't enough.

Descript has strong security, but it's designed for creative teams sharing projects. Scribr is designed for anyone whose notes are under legal or ethical obligation. A therapist's notes aren't the same as a script. A legal interview isn't the same as a video edit.

What we're not trying to be

We're not Otter. Otter won the 'record every meeting automatically' race. Their web player and search are excellent. If your whole team uses Zoom and you want a bot in every room, they're the obvious choice.

We're not Descript. We're not building a video editor or trying to make editing audio as fun as Figma. If your workflow is 'record, edit, publish,' Descript is the answer.

We're not trying to be everyone's tool. We're trying to be the tool you carry with you, the one that gets out of your way, the one that respects what you say but doesn't keep it without asking. The one that works offline because sometimes the best conversations happen where there's no signal.

Android will come eventually, but we're shipping iOS-first because that's where the users we met in testing actually are.

The moment we stopped copying and started listening

That consultant's question, the one about opening your laptop to transcribe a phone call, it wasn't original. It was the same question a researcher asked us. The same question a sales manager asked. The same question from six different jobs that don't look like they'd have anything in common until you realise they're all people whose work lives in conversations, not documents.

Building Scribr meant accepting that we'd never beat Descript at editing. Never beat Otter at enterprise adoption. But we could build something they wouldn't, because they're solving a different problem. They're solving 'how do I record and edit everything.' We're solving 'how do I capture the exact moment when the decision actually happened.'

Three years in, that's still the difference.