Finxan AI – Haru

Inspiration

Finxan AI – Haru was inspired by the limitations of traditional digital learning tools. Most platforms rely on static text or videos, which lack human presence. We wanted to build an AI teacher that feels interactive, expressive, and professional—more like a real instructor than a chatbot.

What it does

Finxan AI – Haru is a voice-enabled AI teacher powered by a 2D animated character. It listens to users, responds using the Gemini API, and delivers explanations through speech, on-screen text, and visual guidance to create a human-like learning experience.

How we built it

We built the system using the Gemini API for intelligent responses, voice input for natural interaction, and a 2D character rendering system for real-time presence. The architecture is modular, allowing seamless upgrades from 2D to 3D in future versions without redesigning the core logic.

Challenges we ran into

We faced challenges in maintaining character stability, synchronizing AI responses with on-screen behavior, handling microphone permissions, and ensuring responses felt professional and original rather than generic.

Accomplishments that we're proud of

We successfully created a stable, voice-interactive AI teacher with a consistent on-screen presence, natural responses, and a scalable design. The system demonstrates that AI teaching can be engaging, expressive, and professional.

What we learned

We learned how to design controlled AI behavior, integrate voice and animation smoothly, reduce response inconsistency, and build systems that balance real-time performance with user experience.

What's next for Finxan AI – Haru

Next, we plan to introduce a fully 3D character, advanced gestures and pointing, richer visual explanations, and enhanced teaching behaviors to further improve realism and learning impact.

Built With

• 2d-character-rendering
• css
• gemini-api
• html
• javascript
• react.js
• vite
• web-speech-api