Inspiration
Managing tasks should feel natural, not like filling out a form. We were inspired by how people talk about their tasks—“Remind me tomorrow,” “I’ll do this later,” “Every Monday at 9.” Twick AI was born to bridge the gap between human language and rigid task managers.
What it does
Twick AI is an AI-powered task and reminder app that lets users create, organize, and manage tasks using natural language—via voice or text. It intelligently understands intent, schedules reminders, supports recurring tasks, works fully offline, and syncs seamlessly across devices when online.
How we built it
We built Twick AI using Flutter for a fast, consistent cross-platform experience. AI-powered natural language understanding—powered by Google Gemini—converts user input into structured, actionable tasks. The app follows an offline-first architecture, storing tasks locally on the device with seamless background synchronization. We used Serverpod as the backend to handle task syncing, database management, and authentication, ensuring reliability and scalability. Serverpod handles the database and migrations, and Serverpod Auth provides JWT-based authentication. Secure Apple and Google Sign-In, data encryption, and iOS widgets were integrated to provide instant task visibility and a smooth, secure user experience.
Challenges we ran into
Accurately parsing natural language into reliable, real-world reminders Designing an offline-first synchronization system that avoids data conflicts Integrating Serverpod end-to-end—everything worked smoothly in local development, but deploying and configuring it on the cloud introduced unexpected challenges.
Accomplishments that we're proud of
Fully functional AI-based task creation using real-world language, powered by Google Gemini Reliable offline usage with seamless background syncing Successfully integrating Serverpod for backend services including task syncing, database management, and authentication—and deploying it to the cloud iOS widget support included in the very first release A clean, modern UI designed for everyday use
What we learned
Offline-first systems require careful planning around data consistency and conflict resolution Natural language UX is more about understanding user intent than matching keywords Simplicity in UI design significantly improves usability and adoption. Building a backend for the first time using Serverpod was a major learning experience, and using Dart on both frontend and backend greatly simplified development and improved productivity
What's next for Twick AI
Wearable and voice assistant integrations for hands-free productivity Public release on both iOS App Store and Google Play Store to reach a wider audience
Log in or sign up for Devpost to join the conversation.