Demo Video:

Hi everyone! I've been working on this platform for a bit to help streamline image reference search and saving, and just released an alpha demo! Please test this out and let me know what you think!https://t.co/a9P6tkWcvM pic.twitter.com/QAFQBfTnVl

— 孟天同 (@fmeng__)April 18, 2023

As a digital artist, I often find myself searching for reference images for my work, spending hours on search engines but still not finding what I need. I wanted to create a tool that would allow me to search for reference images quickly and efficiently, providing automated filters and classification of images based on my search terms. Talking with other artists, I found that this was a common problem, and that many visual illustrator artists struggle to find the perfect reference image for their work. Additionally, sharing and collaborating on reference boards is a commmon need for artists, but currently, there is no easy way to do this.

Artscaper is a tool that solves these problems, allowing artists to search for reference images quickly and efficiently through AI query generation and latent space searching, and share and collaborate on reference boards with other artists. I was in the Startup Hokies Incubator, aiming to grow Artscaper into the best place for artists to save and share references.

The site was all written from scratch in Sveltekit, Typescript, TailwindCSS, Weaviate, GPT-3.5-turbo, Websockets, and .NET Core. We garnered over 50 signups for beta testing in April, and soft launched it with many industry professionals. Unfortunately, this project has been put on hiatus due to other circumstances. I was able to learn a lot about startup development and B2C products, as well as how to scale very reactive and interactive, real-time, and synchronized web applications. Since I was mostly the sole contributor to the project, I also learned the importance of having a cofounder.