About the Project

https://github.com/dhakalaashish/be-better 4 group members

Inspiration

We noticed how often people plan their days mentally but forget to log or review their progress. Writing tasks manually felt tedious, and existing apps were too rigid. We wanted a more human way to track productivity, one that starts with your voice. That’s how BeBetter was born: an AI-powered companion that listens, organizes, and helps you reflect on your daily habits effortlessly.

What it does

BeBetter allows users to verbally log their daily tasks using natural speech.
The app automatically:

Converts speech to text in real time
Categorizes tasks using AI
Tracks progress, completion rates, and productivity trends
Visualizes insights with engaging statistics and charts
Generates personalized summaries to help users improve day by day

It’s like having a personal productivity coach who listens and adapts to your lifestyle.

How we built it

We built BeBetter using:

React + TypeScript + Vite for a fast, modern frontend
Tailwind CSS + shadcn/ui for elegant, responsive design
Supabase (PostgreSQL) for authentication, relational data storage, and API access
Lovable.dev for rapid iteration and cloud deployment
AI speech and NLP APIs for voice-to-text and task classification

Our development process followed a modular approach:

Built the speech input and transcription module.
Designed a task management and visualization dashboard.
Integrated Supabase for data persistence and secure user login.
Added analytics logic for trends, completion rates, and focus insights.

Challenges we ran into

Handling real-time voice transcription reliably with noisy input.
Designing a database schema flexible enough to store dynamic user data.
Ensuring data privacy while using multiple sensors (microphone, location).
Integrating several APIs smoothly within a single frontend architecture.
Keeping the UI responsive and fluid on mobile devices.

Accomplishments that we're proud of

Built a full-stack AI-driven productivity app in a hackathon timeframe.
Integrated speech recognition and task intelligence seamlessly.
Designed an intuitive interface that feels natural and empowering.
Learned to collaborate efficiently with cloud-based tools like Lovable and Supabase.
Presented a concept that balances AI innovation with human-centered design.

What we learned

How to connect AI tools with real-world UX in a meaningful way.
Practical database design for relational apps.
Real-time data synchronization and state management in React.
The importance of privacy and consent when using voice and location data.
Effective teamwork, version control, and communication in high-pressure environments.