Inspiration
Over the past year, generative AI has demonstrated immense potential across various applications. Remodel seeks to enhance the capabilities of these models even further.
What it does
Remodel is a web application that enables users to create custom workflows, with complex input and output formats (image uploads, PDFs, voice recordings, and much more!). We make it super easy to stack multiple AI models in the same workflow. In a matter of seconds, users can create replicable and intuitive applications to suit their personal and professional needs. This was previously impossible with platforms like ChatGPT having only text inputs and outputs.
How we built it
- Typescript + ReactJs + NextJs + Chakra UI for the frontend
- Python + Flask + Supabase for the backend
- Llama 2, Stable Diffusion, Whisper, etc. integrated with Python
Challenges we ran into
- The workflow creation user interface (design + implementation)
- Integrating many complex functionalities on the backend in the given time
Accomplishments that we're proud of
- Managing to build a functional and intuitive UI
- Getting everything working together
What we learned
- Supabase postgres and auth usage + Python and Typescript SDKs
- Advanced typescript usage with database objects
- Getting familiar with the new NextJs 13 code formats
What's next for Remodel
- More advanced input and output features (receiving emails and texts, slack channels etc.)
- API and webhook integrations
- Marketplace for sharing workflows
- Live collaboration for teams
Built With
- flask
- next
- python
- react
- supabase
- typescript


Log in or sign up for Devpost to join the conversation.