FaceDough: Generative AI Label Studio
Mourad Elsheraey · Lead EngineerFaceDough was a Cairo-based synthetic media startup, founded in 2022, targeting the UAE and Gulf markets. The company specialised in high-quality AI-generated visual content: face morphing, image synthesis, and immersive experiences for entertainment, marketing, and social media brands. Mourad designed and delivered a web-based generative AI service for them: end users could select a product, upload their logo, and write a plain-language description, and the platform would then synthesise a set of AI-generated images tailored to their input, ready for campaigns and social content.
The engagement was scoped in three releases: a functional prototype (core upload, prompt, and generation loop), a live demo build put through a formal 14-day User Acceptance Test, and a final hardened public release. End-to-end delivery was estimated at six to nine weeks. Infrastructure was sized to handle 7,200 AI inference requests per month, with six months of post-launch support included. The full project was broken into prepaid and postpaid phases, each gated on acceptance criteria, with a formal WBS sign-off before any change requests could alter scope.
In short: a complete GenAI SaaS product (architecture, backend, API layer, frontend, cloud infrastructure, and UAT) scoped, built, and shipped by one engineer.
A note on the state of the art: What took weeks of scoping, API integration, infrastructure decisions, and phased delivery in 2023 can today be prototyped in an afternoon with tools like Claude Code. A solo developer can now go from prompt to deployed AI image pipeline in a single session, no formal WBS, no phased budget, no six-week timeline. It’s a fair guess that the commoditisation of exactly this kind of capability is part of why FaceDough is no longer operating. The moat disappeared faster than the roadmap.