Hi, I’m Melvin Goxhaj — a Computer Science graduate from Swinburne University of Technology, with a major in Artificial Intelligence, and currently working as a Data Steward / Analyst at ANZ in Melbourne. My focus is on building reliable machine learning systems and models that can be used in production environments for a real world impact.
I’ve had the chance to design and deploy ML applications that has served thousands of users, including StopOdds.com, an end-to-end ML project where I built inference services with FastAPI, set up CI/CD pipelines, added monitoring, and automated retraining. I’ve also explored smart contract security with LLMs, real estate price prediction, and deep learning image classification throughout my time as a student. Most of my work emphasizes simplicity, data quality, and making sure systems can be trusted in real-world settings.
Right now, I’m especially interested in deep learning and reasoning models. I’m always experimenting with new ways to bridge research and production.
Projects
If you’d like to check out what I’ve been working on, my projects are up on GitHub, including:
- StopOdds — production ML pipeline for odds prediction
- AutoScType — type-propagation for smart contract security
- Other experiments in reasoning models, forecasting, and computer vision
I enjoy building things that don’t just work in theory, but actually hold up in practice.
Contact
Feel free to reach out if you’d like to discuss ML systems engineering, deepl learning, or any and all interesting projects: