AI Engineer
& Full Stack Developer
Crafting intelligent solutions and seamless web experiences. From deep learning models to pixel-perfect UIs, I build software that makes an impact.
A blend of cutting-edge AI techniques and modern web technologies to build robust and scalable applications.
Training, fine-tuning and shipping models that hold up outside the notebook.
Type-safe, accessible product surfaces built to scale.
One codebase, native feel, on both platforms.
Fluent across the stack, from tensors to templates.
The infrastructure behind everything above.
A collection of projects that demonstrate my passion for building intelligent systems and beautiful interfaces.
A state-of-the-art biometric identification model trained on a small, heavily unbalanced dataset. It reached 94.7% accuracy — beaten by exactly one published paper, by a single percentage point.
A deep learning pipeline that reads MRI scans and localises tumour regions, turning a slow manual review into an instant, explainable second opinion.
Point the camera at a plate and CalorieFire does the rest. A vision model segments each item, identifies it, and estimates portion size — turning a meal into a calorie and macro breakdown in seconds, with none of the manual logging that makes people quit tracking apps.
A property marketplace where buyers narrow thousands of listings down to the handful worth viewing — location, price band, rooms, features — while agents manage their own inventory behind a role-gated admin view.
An app that teaches Arabic to students, taking them from letterforms to full sentences. Lessons build on stroke-by-stroke script practice, spaced repetition, and pronunciation drills tuned to how the alphabet actually connects.
A classifier that separates AI-generated imagery from the real thing, built to stay accurate as generative models keep moving the goalposts.
An app for people sharing a home. Join a group with your housemates, keep one shopping list everyone can add to in real time, and build a shared recipe list so tomorrow's dinner is settled before anyone opens the fridge.
A consumer marketplace for buying and selling vehicles, built around the filters people actually reach for: make, model, year, mileage and price. Sellers list through a guided flow, buyers compare side by side.
A storefront for digital games with a searchable catalogue, genre and platform filtering, cart and checkout, and a back office for pricing, stock and featured placement.
A family of regression models that turn messy historical data into forecasts teams can actually plan around, with the error bars kept honest.
A platform for car dealerships: inventory management, role-based admin access, and a customer-facing catalogue with deep filtering.
Research prototypes, client builds and experiments that never made the shortlist. Ask me about any of them.
Start a conversationFrom my first line of code to my latest AI model, here's a look at my journey.
After graduating, I've been actively developing a range of projects, from AI models like a Brain Tumor Detector to AI-powered mobile apps like CalorieFire. I am also continuing with freelance projects, currently building a platform for car dealerships.
Pursued my Master's degree, culminating in a state-of-the-art AI thesis on ear recognition. My model achieved 94.7% accuracy on a small, unbalanced dataset, a result surpassed by only a single research paper by a mere 1% difference.
My journey began with a summer internship in July. A few months later, I took on my first freelance web development project, building a solid foundation in the field.
Have a project in mind or just want to say hi? Send me a message.