I'm a tech entrepreneur and AI engineer based in Egypt, and I build products end to end rather than just one piece of them. Over the last few years I've founded and co-founded two AI companies, and both are live, in production, and used by real people, not just prototypes sitting on GitHub.

I'm the founder and tech lead of InnoCanvas, a GenAI SaaS platform I built and run solo. It uses LLM agents to generate a full business model canvas and a panel of AI consultant personas for early-stage founders, plus a scoring engine that rates startup ideas. I built the entire stack myself: the prompts and multi-agent system, the Next.js and FastAPI application, deployment on Supabase and Vercel, and even the billing through Paddle. Since I run it alone, I'm the product manager, the engineer, and the person answering support emails, all at once.

I'm also co-founder and tech lead of Goia, an AI cultural companion app that combines computer vision and AR to help tourists explore Egyptian heritage sites. It's officially approved by the Grand Egyptian Museum and Egypt's Ministry of Tourism and Antiquities, which meant a lot of on-site testing with real tourists at NMEC and GEM before anything shipped. Goia has secured 1.5M EGP in funding and went through Enpact Berlin's incubator and Egypt's GenZ startup program.

Before all this, I studied Artificial Intelligence at Kafr El-Sheikh University, and I also spent time doing graphic design and 3D work freelance, which is part of why product design still matters a lot to me. I don't think of AI engineering and design as separate skills. A model that works but ships in an ugly, confusing interface hasn't really solved the problem.

On the technical side, I work in Python for AI and backend work (FastAPI, data cleaning, computer vision, NLP, machine learning), and I build the actual products in Flutter and Dart for cross-platform mobile, native Android with Kotlin, and Next.js for web. I use Supabase and Vercel to get from idea to production fast, which matters a lot when you're a solo founder and can't afford a six-month build cycle. I recently went through Anthropic's own Claude Code 101, Claude Code in Action, and Introduction to MCP courses, which sharpened how I think about agents and tool use in production AI systems, not just theory.

What I bring beyond the code is a founder's instinct for whether a feature is actually worth building. I've had to make AI features pay for themselves, explain technical trade-offs to non-technical partners including government and museum stakeholders, and mentor other early-stage teams through their own technical decisions. If you need someone who can take a rough idea, or even just a Figma file, and turn it into something real, live, and used, that's the work I do.