Inspiration

Traditional open-source AI video tools like DeepFaceLab are incredibly powerful, but they come with a huge barrier to entry: they require expensive local GPUs and complex Python environment setups. We wanted to democratize pro-level AI video editing so that any creator, regardless of their hardware, could access top-tier rendering capabilities.

What it does

FaceSwap is a fully cloud-powered AI video face-swapping SaaS. Users simply upload their videos via our web browser interface or Android app, and all the heavy lifting is offloaded to our dedicated cloud GPU clusters. It delivers DeepFaceLab-quality results with zero technical setup.

How we built it

We built the frontend web application using Next.js combined with shadcn/ui for a clean, modern interface. The mobile experience is powered by React Native. On the backend, we utilized FastAPI (Python) to efficiently route rendering tasks and manage our cloud GPU infrastructure, ensuring fast and scalable video processing.

Challenges we ran into

Balancing cloud GPU costs while maintaining fast queue processing times for users was a major architectural challenge. We optimized our backend routing to dynamically allocate resources based on user demand.

Built With

  • flutter
  • next.js?fastapi?python?cloud
Share this project:

Updates