-
-
Frontend
-
Before and after
-
Frontend
Inspiration
As a part-time graphic designer, I often struggled with how time-consuming it was to create high-quality product posts and thumbnails - especially when clients needed fast turnarounds. Tools like Photoshop require technical expertise and hours of manual work, which isn’t practical for small businesses or individuals who just want great-looking designs. I realized many content creators, online sellers, and small business owners face the same challenge: they need visually appealing posts but lack the time, skills, or resources to create them. That’s what inspired me to build Fasign — a tool that automates the design process using AI, making professional-looking visuals accessible to everyone.
What it does
Fasign takes a rough product image and a simple user prompt, enhances the prompt using the Gemini API, and then sends the refined version to the Stability API to generate a professional-looking product image. It's ideal for creating eye-catching social media posts, ecommerce listings, or promotional visuals instantly and effortlessly.
How we built it
Fronted : The frontend is built with v0.dev and deployed on Vercel for lightning-fast performance.
Backend : The backend is served via FastAPI on Render, enabling scalable and async API handling.
Google Gemini APIs handle the intelligent prompt transformation.
Stability AI’s inpainting API brings the visual enhancements to life.
Challenges we ran into
Initially, we planned to use the open-source Stable Diffusion SDXL model locally for image generation. While it worked well during development, deploying it for real-time use proved to be a major challenge due to its heavy resource requirements and slow inference on limited hardware. To overcome this, we switched to the Stability AI API, which provided faster and more reliable image generation — but required adjustments in how we handled prompts and output.
Another challenge was integrating the Gemini API for prompt enhancement. We had to ensure that the AI-generated prompts added useful detail and creativity without drifting too far from the user’s original intent.
Accomplishments that we're proud of
Successfully chained together two powerful AI models (Gemini + Stability) in a meaningful way.
Designed a clean, simple UI/UX with React.
Achieved fast and consistent image enhancement results from minimal input.
What we learned
How to integrate multiple AI APIs in a streamlined, user-friendly workflow.
The importance of prompt engineering and enhancement when working with generative models.
Building full-stack apps doesn’t always require traditional backend frameworks simplicity often wins.
What's next for Fasign - Design Eye Catching Posts
Our immediate goal is to focus Fasign specifically on product image enhancement for small businesses. We plan to help users generate ready-to-use social media posts that match their brand theme including color palette, typography style, and layout consistency.
We’re also working on making these posts fully editable so users can:
Change text
Adjust positions of elements
Customize background or layout components All of this without needing graphic design skills or complex tools like Photoshop.
Once the product image-to-post workflow is solid, we plan to expand Fasign’s capabilities to support:
YouTube thumbnails
Web banners
Logos
Professional marketing content
Ultimately, we aim to turn Fasign into a one-stop AI design studio for creators and small businesses who want to build a professional online presence without the usual time or cost barriers.
Log in or sign up for Devpost to join the conversation.