Vorynce's First Month: Building a Voice-First App
Vorynce has been on the App Store for one month. These are our raw reflections — the decisions that shaped the app, the surprises we encountered, and what we are building next.
We launched Vorynce in June 2026, a few months after Foculoom LLC was registered in February. The app we shipped was not the app we originally envisioned. It was better — tighter, more focused, more honest about what on-device AI can and cannot do. Here is what we learned building a voice-first productivity app from scratch, the decisions we made along the way, and where we are going.
The Origin: Why Voice-First
Vorynce started from a simple observation: every productivity app on the App Store is typing-first. Even the ones that support Siri shortcuts or voice input treat voice as an add-on — a second-class capture path bolted onto a keyboard-first interface. We believed voice could be the primary interface, not the backup option. The research backed us up. Voice is three times faster than typing for English input. Speech preserves context that typing strips away. The only missing piece was sufficiently accurate, sufficiently private speech recognition — and Apple's on-device Speech framework, combined with Apple Intelligence, finally made the vision viable.
That conviction shaped every architectural decision. The capture button is the largest element in the app. The task list is secondary to the recording flow. Audio playback and transcription are equal citizens, not a bonus feature. Everything in Vorynce is designed to minimize the time between "I have a thought" and "the thought is captured."
On-Device First, Always
The most consequential decision we made was to commit to on-device processing for everything, not just the Apple Intelligence premium features. Every recording, every transcription, every search index is stored locally. There is no server, no cloud sync (yet), no accounts, no sign-up flow. You download the app, grant microphone permission, and start recording.
This choice simplified our development enormously — no backend to build, no sync conflicts to resolve, no server costs to manage — but it also constrained us. Without a server, we cannot offer cloud backup or cross-device sync. Those features are on the roadmap, but they will arrive only when we can implement them with end-to-end encryption that matches the privacy promise of the core app. We would rather ship late than ship a privacy compromise.
Voice capture: One-button recording with real-time on-device transcription
AI task extraction: Automatic task, deadline, and stakeholder detection (Apple Intelligence devices)
Focus timer: 25-minute Pomodoro with break tracking and capture-during-focus
Soundscapes: Pink noise, nature sounds, coffee-shop ambient, brown noise
Audio journal: Searchable, timestamped, organized by date and project tags
Export: Markdown, JSON, and plain text export for all recordings
What Surprised Us
Three things surprised us after launch. First, the focus timer was the feature that early users mentioned most in reviews — not the voice capture we expected to be the hero. Users wanted structured work sessions as much as they wanted fast capture. The combination of voice capture + timer turned out to be the real product, not either feature alone.
Second, on-device transcription accuracy exceeded our expectations even on older devices. We had benchmarked Apple's Speech framework in our test suite, but real-world usage — with varied accents, background noise, and speaking speeds — showed remarkable robustness. The delta between on-device and cloud transcription on an iPhone 12 was barely perceptible in normal conditions.
Third, the privacy angle resonated in ways we did not fully anticipate. Multiple users told us they had tried cloud-based voice assistants for productivity but stopped because they did not trust their work data on a corporate server. Vorynce's on-device promise — "your voice never leaves your phone" — was the deciding factor for them trying voice capture at all.
Where We Are Going
The first month has validated our core thesis: voice-first, on-device, private productivity is not a niche — it is a category that has been waiting for the technology to catch up. Our public roadmap includes end-to-end encrypted cloud backup, iCloud sync (powered by CloudKit with encryption), Siri shortcut integration for hands-free capture, customizable soundscape mixing, and project-based task grouping with AI-suggested categories.
We are also investing heavily in the accessibility use cases. Voice-first productivity is naturally more accessible for users with motor impairments, visual impairments, or conditions that make typing difficult. Our v1.1 update includes VoiceOver optimizations, dynamic type support, and switch control compatibility — and we are working with accessibility testers to identify the next improvements.
Vorynce was built by a small team with a big conviction: the way we work should not be constrained by the way we type. Voice is faster, more natural, and more private than typing — and the technology to do it right is finally here. Thank you for being part of the first month. The best is ahead.