Meet Dr Sunir Gohil, Principal Data Scientist at Haleon.
By Sunir Gohil
Can you tell us a bit about your background?
My background and journey into data science is rather unique although one which is not unusual for data scientists. I am both a qualified medical doctor and a data scientist. Medical school is normally 6 years and I spent mine at Imperial College London. I had a dream of becoming a surgeon, either in Reconstructive Plastics or Cardiothoracic (open heart surgery). During my med school journey, I spent a lot of time in NYC learning the basics of being a surgeon and how the US healthcare system works. Cardiothoracic and plastics are seen to be the most glamorous of medical jobs, but in reality, this was never the case. Being a surgeon takes long hours, immense dedication, and the drive to want to be the best at your skill, very similar to data scientists!
Once I completed my medical training, I took on a PhD in computer science at Imperial College London. Since then, I have worked in healthcare start-ups developing health chatbots and diagnostic tools. Because my background is in both medical and data science, it gives me a unique perspective on the use of artificial intelligence in health, particularly in consumer health, and how we can change the way we think about health at home.
What is your data science specialty and how long have you been doing it?
At Haleon I focus on a few areas of data science with the teams. Natural language processing is something I find fascinating, which looks to better understand text and language using data science. A single sentence contains a wealth of information, from content and sentiment to intention. Language is difficult to understand, more difficult without context, and almost impossible without any background. Trying to find ways of modeling language so a computer understands, I find extremely interesting especially when it comes to people’s health. How people talk about their health is extremely personal, filled with nuance and emotion, and it is essential that we apply data science techniques in order to get it right! This extends to applications where natural language can be used in other technologies like speech-using devices (eg. Alexa) where interacting with AI becomes even more seamless.
Why did you get into the health sector?
I originally got into health to become a doctor and to learn the skills to become a surgeon. It fascinated me that through effective communication, empathy, and honing a skill, you could heal someone and make them better, and you could see the positive impact you have on someone else’s life just through your actions. Being a medical doctor carries great privilege and honour especially when you can see the difference you make to people’s lives. I soon came to realise that by using the power of data and data science, you can affect many more lives as data science allows you to find new insights, new methodologies, or new understandings that you once could not. I enjoy the role we can play in someone’s everyday health. Before, and since covid, a lot of people looked to the internet, self-help, and over-the-counter to help them heal, or stay healthier for longer and are taking health into their own hands. We’re in a unique position to be able to help many more people on their health journey at the point where they are interacting with their health products, be that vitamins & minerals, or toothpaste.
Are there any areas of DS/ML that you feel will be particularly prominent in the future?
Right now, Generative AI and the use of large language models will drive a lot of the use cases and use of AI in the future. Generative AI can be used in several ways, including the summarisation of complex documents, a better understanding of large data sets, or the generation of content. At its heart, LLMs and generative AI are all about easy-to-use information retrieval for very large data sets. The magic here is anyone can use it to generate information in a way that is easiest for them to understand. This empowers anyone, not just data scientists to leverage power for AI. However, in the future, I envision generative AI and large language models being hyper-up-to-date, pulling on real-time data, rather than currently where data may only be from 2017 or 2021. This means information retrieval will be extremely up-to-date, if not even a prediction of the future.
What’s your top tip for anyone who wants to enter the field?
Coffee. Find something that you are interested in and try to apply the foundations of data and data science to that area. Some people may now know but I am obsessed with coffee. I have recorded the extraction values of all the coffees I have made over the last 2 years in a notebook. This is obsessive but has also given me a great data set to compare coffees, extractions, yields, and TDS and practice my data science skills with more unusual data.
How do you keep up with the field/how do you keep your knowledge current?
Keeping up with current knowledge is essential! I try to mix up the way I keep up to date with a combination of evidence-based papers from Arxiv and other journals, blog posts like Towards Data Science, and YouTube videos. I am an avid coffee enthusiast, so often I like to read about the data science of coffee extraction!
What do you like most about working at Haleon?
First and foremost, Haleon’s purpose to deliver better everyday health with humanity resonates with me as a doctor. This was the reason I started my journey into medical school. In addition, I am surrounded by humble, down-to-earth geniuses in the data science team, and we work together on the combined mission. We have the technical expertise and access to unique data sets that can change the way we think about health at home and the choices we make to stay healthier for longer. Data science and AI can unlock the potential of this vast and interesting data which will change the way health is perceived outside of the doctor’s office in the future.
