Hazem Mathbout

I build AI systems that work in production — not demos, not proofs of concept, but tools people actually use every day.

Over the past 3+ years, I've completed 13 technical projects on Mostaql (100% completion rate, 5-star rated on every reviewed project) and delivered 168+ services on Khamsat, maintaining 5-star ratings with no negative reviews across both platforms.

My work sits at the intersection of two things: turning machine learning models into products people can actually use, and building the Flutter mobile apps that put them in front of users. On the AI side, that's included RAG systems that let companies search their own PDFs and internal documents, a custom LLaMA-based chatbot trained on a client's chemistry reference material, a computer-vision lip-reading application for speech-impaired users built with OpenCV and PyTorch, and a Siamese Bi-LSTM audio model trained to assess Quran recitation accuracy against expert reciters. On the Flutter side, I've built and maintained Facelancer — a freelance-opportunity aggregator running continuously since 2023 — and Hatlak, a multi-vendor e-commerce platform adapted for the Syrian market.

I work end-to-end: Django and FastAPI on the backend, LangChain with OpenAI and Gemini APIs for the AI layer, and Flutter for the interface — deploying and hosting everything myself.

I'm currently open to freelance and fractional AI/Flutter work with startups in Saudi Arabia, UAE, and Qatar, as well as remote full-time roles. If you're trying to get a working AI feature into your product without over-engineering it, let's talk.