JUNE 20, 2024

BRAID’S TEAM

<aside> 💡 Welcome to another article in our series of Tech Team Profiles! Here, we’ll introduce you to the team building Braid’s automated engineering design technology. Thomas Ghorbanian is a Researcher at Braid with a background in physics and computational science.

</aside>

Describe your career thus far.

My studies and work thus far have been at the intersection of physics and computer science. I completed both my bachelor’s and master’s degrees at University College in London. My undergraduate work was focused primarily on particle physics, which eventually led to my master’s work as part of the ATLAS experiment at CERN. As part of the ATLAS project, I used probabilistic deep learning to model potential Higgs Boson events.

I also spent one year studying abroad at the University of Texas in Austin. I joined the Computer Visualization Center at the Oden Institute which was leveraging deep learning to answer questions in material physics. Our research investigated probabilistic deep learning as a means of modeling the relationships between crystal geometry and their physical properties, like band gap and binding energy.

Probabilistic models were the primary focus of my research at university. Essentially, they aim to specify the underlying process that generates observed data. It is an example of unsupervised learning and is sometimes called generative modeling. For instance, in the crystal project, the goal was to learn the underlying distribution of crystal geometries in a way that encodes relevant physics. By sampling from the distribution in specific regions, crystals with desirable properties can be generated.

When I first came to Japan, I worked for a diagnostics and biotechnology company. I discovered the often-overlooked importance of diagnostics compared to therapeutics. Many health problems can be addressed by catching diseases early; I would argue that effective diagnostic tools are also therapeutics, in a sense. As a researcher at my previous company, I worked in computational protein science. More specifically, we were aiming to design and categorize the behavior of different proteins in order to optimize them for different disease targets.

How do you “design” a protein?

Proteins are chains of amino acids linked together. Due to the vast number of possible combinations of amino acids in a protein sequence, there are a large variety of protein geometries and subsequent behaviors. A protein’s geometry heavily influences how it interacts with other proteins and targets of interest. For example, we can use proteins as diagnostic tools, designing them such that they bind to targets in the cell that may indicate disease. Understanding how permutations of amino acids affect the protein’s ability to bind to its target allows us to design better proteins to be used as diagnostic tools.

What originally inspired you to study physics?

Growing up, I was inspired by sci-fi movies like 2001: A Space Odyssey and Interstellar, as well as documentaries by physicist Brian Cox. I was quite interested in the open-ended nature of questions that could be asked on a macroscale about our universe. Over time, I also found the converse to be equally interesting - exploring strange physical phenomena on the microscale, even on the quantum scale. Once I realized that at its core, the discipline provided a principled approach to discover patterns at any scale, it inspired me to pursue it as my field of study, and eventually my career trajectory too.

Thomas hiking the Japanese Alps in Tateyama

Why did you decide to join Braid?

I wanted to be part of a dynamic team that was working on problems in physics and engineering. Additionally, I wanted to use techniques to build disruptive technology for an industry that needs disruption. Manufacturing and engineering design could really benefit from advances in deep learning, high performance computing, and computational design. However, these advances can be difficult to access in an efficient and robust way for individuals and companies that want to leverage them. As a society, we’ve been successful at automating other parts of our lives, but physics still remains quite complex. I am eager to tackle those challenges at Braid.

Can you describe a day in the life of working at Braid?

As a Researcher, I strive to be aware of cutting-edge techniques in our field and scrutinize them. We talk freely amongst the team about what we’re working on and get support from others when needed. At least once a week, we share our results to not just the research team, but the entire company. Working in Shibuya, you are subjected to a sometimes overwhelming number of people when commuting. However, Tokyo food is some of the best in the world, so it’s worth it to have access to so many great restaurants. It’s a great environment to be a part of.

Thomas enjoying a summer festival (matsuri) in Hyogo.

What originally brought you to Japan?