Inspiration

Writing leave requests is a small but stressful task for students and professionals. Many struggle with tone, formality, and clarity, especially when messaging advisors, professors, or managers. Rewriting similar messages repeatedly felt inefficient. LeaveMate was inspired by the idea of turning this everyday friction into a simple, intelligent, and respectful communication experience.

What it does

LeaveMate helps users generate professional, context-aware leave messages in seconds. Users can: Select a leave category and specific reason Choose delivery format (Email or WhatsApp) Generate clear, polite, advisor-friendly messages Save history, manage advisors, and reuse custom reasons Get usage insights and smart warnings for repetitive requests It works offline-first, with Gemini-powered AI enhancing messages when available.

How we built it

LeaveMate is built as a Flutter mobile application with a strong focus on UI/UX and reliability. Frontend: Flutter (Material 3, Dark/Light themes, animations) AI: Google Gemini API, loaded securely via environment variables Architecture: Clean separation of UI, services, and configuration Privacy: No backend, no data upload — everything stored locally Tooling: Built using Antigravity for rapid iteration and polish Gemini is used responsibly to enhance text, not replace user intent.

Challenges we ran into

Achieving perfect Dark Mode consistency across all screens Designing premium animations without causing lag Handling Android intent restrictions (email, browser, sharing) Removing hard-coded API keys and making the app Play Store–safe Balancing AI usage with offline reliability and privacy Each challenge pushed us to improve both the architecture and user experience.

Accomplishments that we're proud of

Fully privacy-first design (offline by default) Secure Gemini integration using environment variables Premium UI with consistent Dark Mode Smart features like custom reasons, advisor management, and analytics A real-world app that solves a genuine daily problem Most importantly, LeaveMate feels production-ready, not just a prototype.

What we learned

AI works best when it supports users, not replaces them UI polish and micro-interactions dramatically improve trust Secure configuration is essential for real-world deployments Offline-first design increases reliability and user confidence Building for users is as important as building for judges

What's next for LeaveMate

Multi-language support Institution-aware templates (college/company specific) Advanced Gemini insights (tone analysis, suggestions) Calendar integration for smarter leave planning Optional cloud sync while preserving privacy-first defaults Web and desktop versions

LeaveMate aims to become a trusted communication companion, not just a message generator.

Built With

Share this project:

Updates