Inspiration
Technology has come a long way: AI can write essays and cars can drive themselves, but waking up hasn’t changed much. Most of us still rely on the same jarring alarm buzz we’ve had for decades. We wanted to fix that. SonmoAI (from the latin word sonmus, meaning sleep) is our answer: a personal sleep companion that helps you fall asleep peacefully and wake up feeling refreshed, by learning your habits and adjusting to what works best for you.
What it does
SonmoAI is designed to make sleep simple and smart. It:
Helps you fall asleep faster with ambient sounds and calming prompts tailored to your mood and routine.
Wakes you up more naturally by choosing the best moment in your sleep cycle to gently rouse you.
Shows your sleep patterns in a way that’s easy to understand, highlighting what helps you rest better.
Offers practical tips like ideal bedtimes, caffeine reminders, or short relaxation exercises, that are personalized for you.
How we built it
We built SonmoAI as a mobile app using React Native with Expo, which let us quickly iterate and test features on both iOS and Android.
AI powers the prompts and helps generate relaxing sounds and morning messages.
Local analysis keeps sleep data private while helping the app learn from your habits.
Supabase manages login and stores your sleep logs securely.
The UI is clean and calming, designed to feel natural and inviting late at night.
Challenges we ran into
Time was a big challenge, building a working AI-driven sleep app in just 6 hours meant we had to focus on the most important features.
Juggling multiple tools, connecting AI and Supabase on mobile devices was tricky under time pressure.
Testing on different devices, making sure everything worked smoothly
Making AI feel human, generating prompts and wake-up messages that felt natural and comforting required a lot of tweaking.
Accomplishments that we're proud of
Built a fully functional AI sleep companion that works without any extra hardware.
Launched a cross-platform app during the hackathon, ready to test on real devices.
Created a calm, intuitive experience that actually helps people sleep better.
Balanced AI innovation with practicality, so every feature serves a clear purpose.
What we learned
We learned how to focus on building features that really matter, even under tight deadlines. We also learned a lot about making AI feel human, helpful, and approachable, not just technical.
What's next for us
Emotion-based bedtime guidance, using mood or journaling inputs.
Smart home integration, like adjusting lights or temperature automatically.
Community features to help users build healthy sleep habits together.
Offline, privacy-first mode, letting the AI run entirely on the device for maximum security.
Log in or sign up for Devpost to join the conversation.