Word Gut

Inspiration

We noticed that most vocabulary apps are rigid and disconnected from real-life usage. They rely on fixed word lists and don’t adapt to how or when people actually want to learn. We wanted to build a tool that integrates language learning into daily routines—naturally, contextually, and with minimal friction.

What it does

• Word Gut helps users learn vocabulary by:

• Letting them extract words from word lists, photos of real-life scenes, or custom text

• Organizing vocabulary by topic and difficulty

• Showing words passively on the lock screen

• Using alarm-based voice interaction: at set times, the app speaks selected words and listens as users repeat them

• Scheduling reviews based on the forgetting curve, optimizing retention

How we built it

We built everything in bolt.new, using:

• TypeScript and React Native for a cross-platform mobile UI

• Expo to speed up testing and deployment

ElevenLabs for:

• Natural-sounding text-to-speech

• Real-time speech-to-text

• Speech-to-speech responses after user input

• A custom memory model to implement spaced repetition

• Alarm + voice interaction flow to actively involve users during daily moments

Challenges we ran into

• Coordinating alarms with seamless voice interaction

• Managing async flows between text-to-speech, speech recognition, and user feedback

• Keeping the user interface minimal while supporting multiple learning entry points

• Ensuring the app feels interruptible and flexible, not demanding

Accomplishments that we're proud of

• Built a complete app in a short time that supports multi-source vocabulary input

• Integrated voice interaction triggered by alarms for better recall and speaking confidence

• Used spaced repetition intelligently to help users remember more with less effort

• Created an app that blends into daily life instead of competing for attention

What we learned

• How to coordinate ElevenLabs services for natural audio interaction

• How to design around passive engagement and interruptible learning moments

• How to structure memory-based review models efficiently

• How much you can achieve in a short time using bolt.new with the right focus

What's next for Word Gut

• Smarter alarm scheduling based on user habits

• Tracking which words were successfully recalled via voice

• Adding native language translations for difficult vocabulary

• Lock-screen interaction upgrades (e.g., tap to hear pronunciation)

• Publishing to mobile app stores and testing with real learners

Built With

• elevenlabs
• expo.io
• reactive.native
• typescript