Professional Experience

At KPMG, I quickly assumed a leadership role on the development team, guiding various projects as we modernised our clients' sales platforms, implemented automation for providing data insights, and accelerated their processes with generative AI where it made sense to do so. I manage junior members, advise on devops processes, and solution data architectures, in addition to undertaking significant pieces of the development work myself.

One of the projects that set me apart was building an LLM wrapper over Salesforce that provided data insights via a chat interface. It was a replica of the generative AI feature that Salesforce themselves released a year later, so it blew away our potential clients when we demonstrated these capabilities long before any of our competitors. Then we of course built the real thing once it was available.

At Moonlight Haptics, we built a device to allow someone with visual impairment to “see” using their sense of touch. In our backyard, we produced an electronic display from scratch, then replaced the pixels with vibrating motors. The intensity of the vibrations corresponded to the distance of the perceived objects, and it was precise enough to distinguish a difference of 10 cm, 1 m away, with 90% accuracy.

What I accomplished at Moonlight Haptics was among my most extraordinary work. You can see below a demo of the lightning-fast rescaling algorithm that feeds to the physical display in real-time, complete with a no-look high five at the end.

As a data scientist, I queried big data, cleaned it, then produced hundreds of visualisations and analyses catered for each of the teams throughout R&D. I built regression models, conducted statistical analysis, and improved manufacturing efficiency and yield. I was particularly adept at automating large swathes of my work to push projects faster while maintaining the high-quality outputs I was known for.

My work also had the potential to be a single point of failure for projects involving the company's biggest customers — it was quite a lot of responsibility!

Qubits, naturally, are most famous from the world of quantum computing — you may have seen the intricate golden machines built at IBM or Google, but these computers operate at temperatures colder than outer space. The qubits I worked with at Princeton were room-temperature, which is one of the reasons they could be used to precisely locate and identify individual atoms of biological molecules that would otherwise freeze.

Although engrossed deeply in the world of physics here, I was no theorist. I was, however, able to write hyper-efficient code to conduct tens of billions of automated experiments in a matter of months, then generate insights that let us learn much about how our qubits interacted with each other and their environment. This led to my co-authorship on a paper which we later published to PRX Quantum.

Your colour printer, if you've ever noticed, has three ink colours: cyan, magenta, and yellow (and often black as a fourth). Mixing them in different amounts produces any desired colour on the page. But what if you could make an ink that changes between those three colours when you give it an electric signal? It would allow you to make the screen on your mobile phone look like ink on paper while it plays video. It would be gentle on the eyes at night and vibrant outside in direct sunlight.

Working at Halion Displays was what inspired me to eventually pursue Moonlight Haptics a few years later. The goal is brutally difficult to reach, you're forced to push the limits of human knowledge in disparate fields you were never qualified to pursue in the first place, but there's a small chance you can make something that can really change our lives for the better, that only you were brave enough to try.