engage.ai

• 

  Landing page

• 

  Auth page

• 

  Dashboard page

• 

  Exercises page

• 

  Intervie page

• 

  Interview page

• 

  Profile page

• 

  Exercises voice page

Inspiration

I took the inspiration from Duolingo. They help people to learn languages and I thought helping people to improve their speech would be something AI could help with. I held this idea in mind for a while, but when randomly the hackathon popped and I saw Tavus as a partner, everything clicked for me. Their implementation looked like everything I wanted and needed to make my idea come to life. The goal of EngageAI is to bridge communication gaps and encourage deeper engagement in every interaction and this is what I put into the name behind the app.

What it does

We provide a realtime AI agents (video personas) backed by Tavus.io, that listen and analyse user's live speech and provides a helpful insights about what can be improved.

How we built it

Code was written fully by bolt.new. Tools I've used: bolt.new, tavus.io, supabase, google oauth, netlify, gemini 2.5 pro for Tavus personas prompts.

Challenges we ran into

Sometimes bolt hallucinates, reimporting helps, but that means redeployment of the core code. I think it's a browser and cookies thing, but preview was absent sometimes.

Accomplishments that we're proud of

Having a working product, at the current stage I consider it to be mvp.

What we learned

Bolt can be huge for creating the applications. But that also means to achieve the best results you would need to utilize it as much as possible.

What's next for engage.ai

Reaching first paying customers. We'll see if people enjoy using the product and we can get our pfm.

Built With

• bolt
• javascript
• netlify
• react
• supabase
• tavus