QuitMate

QuitMate is your AI-powered quit smoking coach — using CBT, video, and voice to retrain your brain and help you quit in just 6 days.

Comment

Inspiration

QuitMate was born from something deeply personal. After years of failed attempts, our founder finally quit smoking — not through willpower, but by applying a structured method: blending CBT techniques, self-reflection, and daily psychological rewiring. It wasn’t a product idea at first — it was survival. And it worked. That same method was then shared with friends and family — and it worked for them too. That’s when the realization hit: what if this could help thousands — or millions — if powered by AI? The Bolt Hackathon gave us the perfect opportunity to bring that vision to life — fast, focused, and with the right tools to make it real.

What it does

QuitMate is your personal AI quit coach — designed to help you quit smoking in just 6 days. It delivers a science-backed daily program powered by:

  • Conversational AI for personalized, judgment-free coaching
  • CBT-based activities to break cravings, triggers, and mental loops
  • Multilingual AI coaching that speaks your native language — via voice and video

How we built it

We used Bolt.new to rapidly build and iterate on the QuitMate MVP — combining multiple AI tools into one seamless experience.

  1. Frontend: Built with React using Bolt’s UI kits to design a clean, mobile-first interface optimized for daily interaction
  2. AI Video Coach: Integrated with Tavus to deliver real-time, daily video interactions — creating the feeling of a personal coach who checks in face-to-face 3.Voice Coach: Used ElevenLabs for multilingual voice conversations — enabling the coach to literally “speak your language” in a natural, empathetic tone
  3. Conversational AI: Leveraged the conversational capabilities of Tavus and ElevenLabs to deliver dynamic, voice/video-based dialogue — not just scripted content
  4. Behavioral Engine: Designed and implemented daily CBT-based prompts, reflection flows, and mental rewiring exercises based on proven smoking cessation psychology

Challenges we ran into

We faced several challenges while integrating Tavus and ElevenLabs APIs. Early on, we encountered frequent API errors and unclear responses that made debugging difficult. However, with the help of their documentation, community forums, and AI-assisted debugging, we were able to isolate the issues and resolve them. These roadblocks forced us to better understand how each service handles requests, media processing, and error handling — ultimately helping us build a more stable and reliable system.

Accomplishments that we're proud of

  • Built a fully functional AI-powered quit-smoking coach in just a few days — combining GPT, Tavus, and ElevenLabs into a cohesive experience

  • Integrated multimodal AI (video, voice, and conversation) into a structured, behavior-change program based on CBT principles

  • Created a complete 6-day journey with personalized, emotionally aware coaching content designed to rewire user habits

  • Designed a user-friendly, mobile-first interface using Bolt — fast, intuitive, and built for people going through a vulnerable, life-changing process

  • Turned a personal quitting story into a scalable product, ready to help thousands — and built it in 10-15 days

    What we learned

    Building QuitMate showed us just how fast and enjoyable app development has become. With tools like Bolt, we went from idea to a fully functional AI-powered product in days — something that used to take weeks. We learned that prompting is everything — it’s not only how we built the product, but how the product itself delivers value to users. Integrating tools like Tavus and ElevenLabs wasn’t always smooth, but it pushed us to get better at debugging APIs, working with emerging tech, and crafting a cohesive user experience across voice, video, and conversation. But more than anything, turning a personal quitting journey into a real, professional product gave us the belief that we can build solutions that truly help people — and the confidence to keep going.

What's next for QuitMate

  1. Launch a public beta to gather real-world feedback from people trying to quit smoking
  2. Add real-time craving tracking to offer immediate support when users feel triggered
  3. Introduce rewards and gamification — including on-chain badges and milestones for staying smoke-free
  4. Build a supportive community layer, where users can share progress, tips, and encouragement
  5. Partner with therapists and smoking cessation experts to refine and validate the CBT-based coaching flow
  6. Make it a commercial product

Built With

  • elevenlabs
  • netlify
  • react
  • supabase
  • tailwind
  • travus
Share this project:

Updates