<p>We are 2 ML Engineer with strong experience in Diffusion models and ComfyUI. We noticed you could build state of the art image workflows with open source models but it was too hard for companies (brands, marketing agencies, creation studios...) to build on these technologies.</p><p> We decided to build an "access layer" to open source AI models with a focus on image diffusion models. It was our first shot at this: we developed custom workflows & models and made them available through our web interface. </p><p>Our project is a web based platform where users can use advanced AI modules to edit images (upscale, restyle...).<br><br>It runs ComfyUI workflows and serve a simple user interface where non technical users can play with the workflow parameters and inputs.<br><br><b>It supports:</b><br>- Login (Supabase);<br>- Payment/ Subscription (Stripe);<br>- User generation history (supabase + blob storage);<br>- Workflows run on ComfyDeploy;<br>- Front stack is Svelte, Sveltekit &amp; Vercel.</p><br><br/> ✅ $500 in ARR ✅ 2 customers ✅ Business model: SaaS ✅ Built with Supabase, Vercel, TypeScript