SELECTED WORK
I build ML-backed platforms, publish research, and ship production-ready web products — from demand forecasting pipelines and DPO fine-tuning to React frontends and browser-based CPU simulators.
Research: IEEE IICC 2026 · DPO + QLoRA
Stack: React · FastAPI · PyTorch · TypeScript
Ships: Production-deployed, zero fluff
Featured Project Gallery
ABOUT ME
I'm Ali Badawi — an IEEE-published engineer and full-stack developer. I care about structure, clarity, and building things that are both technically rigorous and genuinely usable, from ML pipelines to production web products.
Current Focus
I research and build ML-backed systems, ship production web products, and collaborate with founders and teams on launch-ready experiences — from fine-tuning language models to frontend architecture.
Co-author on an IEEE-published paper (IICC 2026) on efficient summarization with small language models via DPO fine-tuning.
How I Work
I start with the problem — whether that's a model architecture decision or a UX hierarchy question — then move into implementation with an emphasis on rigorous, maintainable code and clear communication.
What You Can Expect
Core Stack
RESEARCH
Co-authored with Y. Tamer and M. Bahgat. Presented at ICEENG, Cairo, May 2026.
Qwen3-0.6B fine-tuned with DPO + QLoRA on a single RTX 4090 in ~3 hours, outperforming supervised fine-tuning on both automatic quality and factual consistency metrics.
EXPERIENCE
EfficaTech Egypt · Cairo
Aim Technologies · Cairo
TECH STACK
Frontend
Backend
AI + ML
Tooling
Featured Project