MindPod

• 

  MindPod lets you create and experience AI-generated tales, blending creativity and interaction in one place.

• 

  it lets you chat with a particular document and also option to create a notes

• 

  Notes generated here can be modified and downloaded as per need

• 

  Coding environment which helps to write a code and add it to notes with explanation

• 

  A game space to relax

Inspiration

We wanted to create an AI that truly respects privacy. Most AI tools depend on cloud models, but we imagined a companion that runs entirely offline, learns from your notes, helps you study, and never sends data anywhere. That idea became MindPod — your personal, on-device AI companion.

What it does

MindPod lets users: Upload documents and chat with them locally. Take AI-assisted notes and summaries. Track study progress and daily tasks. Practice with quizzes generated from uploaded materials. Interact fully through voice commands for a natural experience. All processing happens directly in the browser, powered by Gemini Nano (on-device AI) — no internet, no servers, no data leaks.

How we built it

We built MindPod using: Chrome’s on-device AI APIs (Prompt API + LanguageModel) WebAssembly (WASM) for multi-language code execution (C, C++, Python, JS) IocalStorage for local data storage PWA architecture for full offline use React (TypeScript + TSX) for a modular, interactive UI

Challenges we ran into

Integrating Gemini Nano to work completely offline across devices. Handling voice input/output reliably in browser contexts. making these voice inputs working locally and also in story tab the story narration. Managing prompt streaming and state transitions in React. Designing an intuitive interface that feels human yet minimal.

Accomplishments that we're proud of

Built a fully offline AI system powered by the browser’s native AI engine. Designed a voice-first interface that’s natural and distraction-free. Created a secure, privacy-first learning companion. Integrated multi-language code execution with WASM — all locally.

What we learned

We learned how to use the Chrome Prompt API and LanguageModel for on-device reasoning. Most importantly, that AI doesn’t need the cloud to be useful — it can live completely on your device.

What's next for MindPod

Add adaptive learning and memory to track mastery over time.Enable speech feedback for natural conversation loops.Introduce AI-driven learning paths and progress analytics. Package MindPod as an installable offline PWA for schools and learners worldwide.

Built With

• gemini-nano
• react
• tailwind-css
• typescript
• wasm
• web-speech