Squinia

• 

  Homepage

• 

  Learning Streaks

• 

  Achievements

• 

  Learner Dashboard

• 

  Scenarios Examples

• 

  Demo

Inspiration

I have experienced the problem myself and I have seen it in my other peers, we have the skills to a job, but when is time to for us to communicate clearly, and professional in interviews, presentations, or workplace discussions we struggle, that was why i built Squinia, to help us practice communications and get instant structured feedback on where exactly we need to improve and we rinse and repeat.

What it does

Squinia helps people get better at communicating professionally by using realistic chat, voice, and video simulations.

How we built it

We used LLM APIs for AI inference to chat, give feedback & create scenarios, WebRTC from Livekit, with Google voice, Deepgram, & Elevenlabs for realtime voice communication, and then NextJs for frontend and FastAPI for backend.

Challenges we ran into

Making the voice communication realtime, we started with websocket, but it couldn't give us a realtime communication so i had to use WebRTC from Livekit.

Accomplishments that we're proud of

Making the voice communication realistic.

What we learned

Talking to customers and validating the product, building a community around it, and able to have a great AI project to help people get better.

What's next for Squinia

Squinia will be the go to platform for people to learn professional communication, with great community of learners and human professional coaches too to give more insights while people practice their insights in Squinia too.

Built With

• fastapi
• gcp
• gemini
• javascript
• llm
• nextjs
• python