Khaled saad abbass abdelmaged

I build the layer between "the model works" and "the model runs on a $150 phone." 6 years of software engineering across mobile, backend, and on-device ML. At BOT Company, I design OCR/document-scanning pipelines — most recently a 2-head multihead CTC architecture handling mixed Arabic text and numerals with sub-pixel accuracy on constrained hardware. Shipped production apps (Face Auth — live on Google Play and App Store), open-source Flutter architecture packages, and camera-to-GPU pipelines built as C++ NDK plugins. Currently finishing an MSc focused on mobile AI, actively looking for remote or contract work in edge ML, computer vision, or systems-level mobile engineering. If your team needs someone who can go from research paper to running inference on-device without a "we'll optimize it later" — that's the gap I fill.