Long story short
I designed a mobile app that turns raw video into ready-to-publish clips with automatic captions. The whole point was to make subtitling feel effortless for creators - so I streamlined the workflow, layered in AI speech-to-text, and built a simple editor for styling and timing. The payoff: captions that used to take ages by hand now come together in minutes.
Problem
Captioning is one of the most repetitive parts of publishing, and the existing tools are desktop-first and fiddly - creators needed to upload, caption, fix, style, and publish entirely from an iPhone, with AI that’s instant but still under their control.
Solution
I designed an iOS-first app built around one-hand, gesture-based use - captions edited live on the video, fully editable AI speech-to-text, a minimal interface for heavy text editing, and onboarding tuned for instant value.
Result
Designed and shipped to the App Store in just 3 weeks, landing a 4.6★ rating with stable in-app purchases and clear early-user validation - captions in minutes instead of manual work.
Context
Capsula AI is a mobile-first AI app that lets content creators generate and style video captions right on their iPhone.
It started from nothing but an idea - and was designed and shipped to the App Store in 3 weeks.
The bet was simple: for modern creators the phone is the main tool, yet most captioning apps are still desktop-first or overcomplicated.
So the goal was instant value - without taking control of quality or style away from the creator.
The core hypothesis was that for modern creators, the smartphone is the primary working tool, while most captioning solutions remain desktop-first or overly complex.
The goal was to deliver instant value while keeping creators in full control of quality and visual style.
! Problem
The promise was instant AI captioning, but the real test was making AI feel genuinely useful on mobile - fast, accurate, and trustworthy in a real creator’s hands.
Creators need captions fast, yet captioning is still one of the most repetitive parts of publishing a video.
It all had to fit a mobile-first flow: upload a video, generate captions, fix mistakes, tweak the style, and publish - entirely from an iPhone.
And the AI couldn’t become a black box. People wanted instant results, but also clear control over text, timing, and look.
Design challenges
Three things made this one tricky:
A brutal timeline
Full design cycle and App Store launch in 3 weeks.
Mobile + AI constraints
Complex flows, text, and timing on a small screen - and the AI output had to be fast, accurate, and editable.
Solo designer
UX, UI, flows, caption-editing logic, motion, onboarding, and App Store assets - all mine.
Solutions
What was done and what benefits were achieved for each solution:
iOS-first
Designed around:
bottom navigation
one-hand use
gesture-based interactions
Live caption preview
Captions are edited right on top of the video - no context switching, far less friction.
Editable AI output
The AI gives instant results, and you stay in charge:
fix recognition errors
adjust the text
restyle captions in real time
Minimal UI for a heavy task
A deliberately stripped-back interface keeps the focus on content while handling text-heavy editing on a small screen.
Onboarding built for fast value
A lightweight onboarding flow gets first-time friction out of the way so people feel the value immediately.
Research, Discovery & Validation
Capsula AI was rolled out in October 2024. For discovery and research, I used different techniques:
Customer reviews & CustDevs. User tests, quick corridor checks, and customer interviews.
Prototype / MVP test. Feedback from real users through support and product analytics.
Competitor analysis. Combed App Store apps to spot the common patterns - and the gaps.
Impact & Results
Outcomes
4.6 rating on the Apple App Store.
Stable in-app purchases.
Clear early user interest and validation.
Reflections and Takeaways
Designed and shipped a mobile AI caption editor from idea to App Store in just 3 weeks, as the only designer.
Coming from a B2B SaaS world, this was a deliberate jump into consumer mobile UX, where:
speed matters more than configurability
the UI should disappear, not explain itself
AI only earns its place when it delivers instantly and still leaves you in control
It sharpened my mobile-first, AI-driven product instincts a lot.
