Mason Cusack

Here I earned a first class honours degree in Computer Science, graduating as the top student in my cohort. During my time at Birmingham I conducted an undergraduate research project implementing a deep learning-based computer vision system for handwritten mathematical expression recognition, won The Birmingham Project with IBM, Capgemini Community Challenge, and the university's own Emerging Technology Competition.

I started my career building apps at an ed tech and robotics startup, funding the opportunity through my university's bursary programs for international work experience and underprivileged students. I coded Android apps in Java, learned about the maker movement and hardware hacking, learned phrasebook Mandarin (improved somewhat since), and made new friends from all over the world.

I was a Software Engineer and Applied Scientist at Microsoft during and after university. After exploring various technologies, I specialised in custom machine learning applications, training and deploying ML models to solve real world problems - typically involving Computer Vision or NLP. The variety of projects I worked on allowed me to travel as far as Africa, The US, and across Europe, embedding with major companies and governments to help them utilise AI in software. Here I learned through experience effective methodologies for productive data science-centric software development.

At ASOS I worked on high scale, real-time personalisation, developing and serving deep learning models for recommendations and search. My team drove millions in incremental revenue with these projects. I mentored MSc and PhD students from University College London to publish research on the use of graph neural networks for returns prediction, was heavily involved in hiring and mentoring, and was on support for our deployments serving thousands of requests per second, each month and during peak periods like Black Friday.

For much of 2023 I worked independently on my first SaaS project, Relevant, while intensively studying and practicing Mandarin in Taiwan. This project introduced me to new tools and technologies including JavaScript, React, Supabase, Stripe, MongoDB, and Railway.

Now I continuously build and release Micro-SaaS products, such as Watch Craving, through my company Relevant Software.

Marshmallow is a unicorn British fintech startup. Yep, another one. 🇬🇧

My job is to run and develop Marshmallow’s machine learning platform - enabling, serving, and scaling applications like fraud detection and dynamic pricing, while building and managing a small, ambitious team.

Get in touch with opportunities or ideas for collaboration. Check out the following links for contact or more information.

• Download CV