Inspiration
Our project was driven by the challenge of making design tools accessible and effective for users at all skill levels. Many amateur designers struggle with creating professional-quality designs and feel overwhelmed by complex design software. We sought to address this gap by integrating AI with Canva's user-friendly design tools. Our goal was to empower users to enhance their designs effortlessly, regardless of their design experience.
What it does
The app leverages AI to assist users in refining their Canva designs. It enables users to:
- Fetch a list of their Canva designs and receive tailored suggestions for improvements.
- Get recommendations on fonts, colors, and overall design elements.
- Receive dynamic feedback directly within the Canva environment, making the design process more intuitive and less daunting.
- Create comments on your design based on the previous generated feedback.
- Live preview of your design using Augmented Reality anywhere you want.
- Export to different formats pdf, gif, png, jpg.
How we built it
The frontend of the app is designed for mobile devices, following Canva’s Developers Design Guidelines to ensure a seamless user experience. We used a combination of modern technologies:
- Frontend: Swift for the mobile interface, with a focus on creating a responsive and intuitive design.
- Backend: Python for handling interactions with OpenAI APIs, providing the AI-driven suggestions.
- API Integration: The core of our project revolves around Canva’s Connect API, which allows us to fetch user designs and integrate real-time AI feedback.
Challenges we ran into
Integrating the Canva Connect API was a significant challenge due to the complexity of handling various design formats and ensuring real-time feedback. Balancing the AI's suggestions with the user’s original ideas required fine-tuning and multiple iterations to achieve an optimal result. Additionally, ensuring a smooth and responsive user experience across different devices posed some technical hurdles. Additionally, implementing Augmented Reality to have a live preview of the design in the real live was a challenge because we had to play with the design dimensions and the space of the environment in order to create realistic proportions.
Accomplishments that we're proud of
We’re proud of successfully integrating Canva’s Connect API to create a tool that enhances the design process. The app's ability to provide meaningful AI-driven suggestions while maintaining an intuitive user interface is a major accomplishment. We also take pride in adhering to Canva’s design guidelines, resulting in a cohesive and user-friendly experience. Additionally, giving the capability of live previewing the design with printing it was really awesome. Who doesn’t want to see their design on a billboard?
What we learned
We learned the importance of balancing technology with user experience. Effective AI integration can significantly enhance a tool, but it’s crucial to ensure that it complements and amplifies the user's creativity rather than overshadowing it. We also gained valuable experience in working with the Canva Connect API and handling complex data interactions.
What's next for our application
We plan to expand the app’s features by adding more AI-driven capabilities and refining the user experience further. Future updates will include enhancements based on user feedback (using RAG capabilities) and leverage the Connect API with the upcoming new features. Our aim is to continue empowering users and making design accessible and enjoyable for everyone.
Improvements:
- Comments api is not rendering comments in the Canva UI - List designs was sometimes returning a continue attribute - Get design could return the comments too, there is no way to get the comments from a design unless I know the ID (which is impossible through the api)
- There is no way to get the list of all the folders through the connect api
- Requesting access for the Autofill API was not granted (even after trying to request access through the for twice and getting in touch with support )
Built With
- connectapi
- kubernetes
- langchain
- openai
- python
- swift
Log in or sign up for Devpost to join the conversation.