-
Landing page
-
Full Sign up/Login with Auth logic
-
User Dashboard
-
Image upload page
-
User practice page
-
Logo
Inspiration
We were inspired to build this app because we wanted to help adolescents and young adults improve their handwriting. Impressions matter a lot for young adults, whether it's teachers, professors, or members of your team at work. Neat handwriting is often linked to positive qualities like carefulness, attention to detail, and a sense of organization. Unfortunately, many people are never taught proper handwriting techniques at a younger age. While this may not matter much when typing online, it can become a real obstacle in situations that require pen-and-paper writing. Improving handwriting is also difficult without knowing exactly what aspects need work—whether it’s spacing, sizing, or alignment. Even small improvements in these areas can make handwriting more legible and polished.
What we learned
We learned how to collaborate effectively in a hackathon setting, trying our best to pick up unfamiliar technologies, and overcome technical hurdles under time pressure. We also gained valuable insights into the complexity of training machine learning models, and the importance of balancing custom development with existing tools.
How we built it
We built the project with a React and Next.js frontend styled with Tailwind CSS, and a backend powered by FastAPI, OpenAI API, and Supabase. To streamline development and deployment, we containerized both the frontend and backend using Docker and managed them together with Docker Compose.
Challenges we ran into
One of our toughest challenges was configuring Docker for both the frontend and backend services and ensuring they worked seamlessly together. An even bigger challenge came when we attempted to train our own handwriting recognition model rather than relying solely on existing OpenAI models. Our custom model struggled with accuracy, averaging only around 30% in letter recognition during testing.
What it does
LetterBuddy provides personalized AI feedback to help users identify weak points in their handwriting and improve through guided practice. It focuses on details such as spacing, sizing, and alignment to help users develop neat, legible handwriting. As a result, adolescents and young adults can feel confident in their writing etiquette and make a good impression in an academic setting.
Accomplishments that we're proud of
For most of us, this was our first hackathon experience, time seemed to fly by and progress came slow. When it came to our tech stack, it was also our first time working with several of the technologies we used. Despite the steep learning curve, we successfully built a working app and gained hands-on experience with tools like Docker, FastAPI, and Supabase.
What's next for LetterBuddy
We plan to train our own model specializing in specific handwriting recognition to have better accuracy. Our model should recognize letters, spacing, and alignment with high confidence. We also plan to extend LetterBuddy to tablet users on Apple and Android, to then have compatibility with Apple Pens and other ePen writing tools. This mobile compatibility will then come with real time feedback on handwriting during practice sessions.
Log in or sign up for Devpost to join the conversation.