The first product decision I made when starting Bonfiyah was that there would be no web app. Not now. Not after we hit a hundred thousand users. Not when an enterprise customer asks. Not ever.
The second product decision I made, much harder, was to keep saying so out loud, in public, when I knew it was costing us users. Because the request comes in weekly. "I love the product but my team is on Windows." "I'd pay double if there was a web version for Chromebook." "My CISO won't approve a personal-iPhone app."
The answer is no. Here is why.
The architectural reason.
Bonfiyah is built around features that web browsers structurally cannot ship. Lock-screen Live Activity. Dynamic Island. Native iOS Calendar and Contacts integration. On-device live captioning via Apple Speech (SFSpeechRecognizer). On-device voice signatures running on Apple's on-device AI. The recording-status widget on the home screen.
None of this exists on the web. None of it will exist on the web. A "Bonfiyah web app" is a Bonfiyah without 40% of what makes it Bonfiyah. The features that compete with Otter and Granola on a Mac are downstream of the iOS-native primitives — and once you start building two products, one for the platform and one for browsers, you are no longer making the same product.
The privacy reason.
The privacy posture is structurally tied to the platform. Audio capture happens on your iPhone. Transcription happens on your iPhone (with cloud assist via an upstream transcription provider for the heavy multilingual lifts, and the data leaves only when you've consented to that path). The voice library lives in your private iCloud, scoped to your iCloud-bound device cohort. None of this is achievable from a browser without re-architecting the whole stack around a server we'd have to operate, secure, and trust.
That's not a fixable problem. The browser is a different threat model. If we shipped a web app, we'd be shipping a different privacy product, with a different attack surface, a different consent story, a different relationship to your data. We'd have to either water down the iOS posture to match the web reality, or maintain two stories and quietly hope nobody notices the gap. Both options are worse than the option of not building it.
The strategic reason.
Bonfiyah keeps a deliberately bounded surface area — one Swift codebase, three platforms via Catalyst, no platform-specific quirks to maintain across React/Vue/native iOS/native Android/Electron at once. A web app would, in slow motion, become a second codebase. Then a second team. That's a SaaS playbook Bonfiyah is specifically not running.
The other strategic reason: native iOS is becoming a moat. The post-everywhere-app fatigue is real; users are increasingly choosing native experiences over web wrappers when both exist. Otter, Notta, and Fireflies all started with web-first or web-equal stacks; their native iOS stories are perpetually catching up. Bonfiyah's iOS-deep strategy is not a constraint we wish we could shed — it is the reason we can ship Live Activity and Dynamic Island and on-device voice ID at all, and it is the reason we don't look like the other apps in the category.
What we'll do instead.
If you need to share a Bonfiyah artefact with a Windows colleague, we ship branded PDFs that read perfectly in a browser, with the Email Intelligence deep-link scheme baked in for the people who do have the app. If you want the AI Summary in someone else's inbox, the email export is the canonical render — your colleague reads it in Gmail or Outlook, and the deep links degrade gracefully to a public web view if they don't have the app installed. We meet the web at the export layer, where it makes sense to meet it, and not in the recording layer, where it doesn't.
If your team is uniformly Windows or you need a browser-first product, you should not use Bonfiyah. We are not the right tool for that situation, and I'd rather you went elsewhere happy than landed here disappointed. There are good products in this space — Granola is excellent on Mac, Fireflies is mature for revenue teams, Otter has the largest customer base. Pick one of those.
For everyone else: the no-web-app rule is not a temporary state. It is the strategy.
— Richard
