Why we built Scribr instead of copying Otter
Last spring, a therapist emailed us: 'I need to record sessions on my phone, but I'm not uploading them to some server I don't control.' That single message shaped how we thought about Scribr. It wasn't about beating Otter at their own game. It was about building something fundamentally different.
The mobile-first assumption nobody else made
Most transcription tools treat the phone as an afterthought. You record on mobile, then upload to the web to do anything useful. It's a tax. It's friction. It exists because the original products were designed for desktop users who happened to need a mobile companion.
We started with the opposite assumption: knowledge workers live on their phones. A consultant takes notes between meetings in the car. A researcher records observations in the field. A salesperson jots down a call recap while walking back to the office. None of these moments happen at a desk.
So we built Scribr as a mobile app first. The Quick Record widget lives on your home screen. Siri can start recording with a voice command. You can transcribe audio right there on the device, without ever opening a settings menu or waiting for a server. That privacy layer - Whisper and Apple Speech running locally - wasn't a feature we added later. It was the premise.
Privacy as a starting point, not an afterthought
The therapist's email stuck with me. Healthcare workers, lawyers, researchers - these are people whose conversations contain sensitive information. They need to take notes, but they shouldn't have to bet their practice on a third-party server's security posture.
Our Free tier does exactly this. On-device transcription. No upload. No cloud sync. No opt-out. The audio never leaves your phone. If that's all you need, you pay nothing and your work stays yours.
When you want cloud transcription - because sometimes you need to record a longer call or share notes with a team - you move to Pro. That's when we use Deepgram. But it's your choice. You know what's happening. And if you're handling regulated data, our GDPR Compliance Modes and Vault Mode (AES-GCM encrypted notes) are built in, not bolted on.
The contrast with web-first tools is sharp. They assume cloud from day one. Privacy is a feature you pay extra for, if it exists at all.
What happens when you don't force a calendar integration
One design choice we didn't make: we didn't build Scribr as a meeting bot that syncs with your calendar. No Zoom plugin. No Google Calendar hook.
Why? Because most knowledge workers don't record every meeting. A lawyer records witness interviews and client calls, not internal status updates. A consultant records discovery sessions and contract walkthroughs. A researcher records interviews, not team standups. And many of these conversations happen outside formal calendar blocks.
We built Scribr around what actually happens: you open the app when you need it. Record. Transcribe. Extract action items if you want them. Move on. No dependency on your calendar tool. No permissions needed from your IT team. No bot notifications cluttering your chat channels.
For sales teams and larger groups where Contact Intelligence and Note Sharing matter, the Team plan adds that capability. But we didn't make it mandatory. Not every user wants their notes stored in the cloud or shared across a team.
The action items thing is quietly useful
We added AI summaries and action-item extraction fairly early, once cloud transcription was stable. Not because it was trendy, but because of how people actually used the app.
After a call, users were scrolling through transcripts, manually highlighting what they needed to do. 'Follow up with Sarah on the budget.' 'Send contracts to legal.' It was tedious. Repetitive. Exactly the kind of friction that slows down knowledge work.
So we built the Action Items Widget. It pulls out tasks from your notes and surfaces them on your home screen. You can trigger it via Siri too. For Pro users, we also send push notifications when action items are ready, so you don't have to remember to check.
It's not magic. It's a time-saver. The kind of thing that compounds over hundreds of calls and conversations.
Why we don't pretend to be something we're not
We're not a meeting bot. We're not a web-first tool with a mobile skin. We don't require your calendar, your email, or your organisation's trust to function. We don't force you to choose between privacy and convenience.
We're a tool for people who record conversations on their phones, need those conversations transcribed accurately and privately by default, and want to extract value from their notes without jumping through hoops or surrendering their data.
If you need something else - calendar integration, team automations, a web-first experience - there are tools built for that. We're not those tools. We're built for the knowledge worker who walks out of a meeting and needs to capture what happened. Right then. On the device they're already holding.
The therapist who prompted this whole direction is still using Scribr. She's never uploaded her session notes to our servers. She pays nothing. That's the product working as intended. What would your relationship with a transcription tool look like if privacy wasn't a negotiation?