February 11 marks the seventh International Day of Women and Girls in Science, a global event that heralds the importance of gender balance in the fields of Science, Technology, Engineering, Art, and Math (STEAM). To show our support and get everyone talking, thinking, and learning about women and girls in science, we caught up (virtually!) with Shveta Gupta, Data Science Manager at NSW Department of Customer Service (DCS) on inspiring girls and women to consider studies and careers in the growing field of information and communication technologies, as well as the important work Shveta and her team are doing across NSW government.
Hi Shveta! Tell us about yourself? What is your role at the Department of Customer Service?
I have 15+ years of experience in data science and started my career within the information technology space. Six years ago, I moved into more of a data analytics and data science space and since then, I've pretty much been working in the business and data side of things, specifically within NSW government.
I've been a part of the Data Analytics Centre at DCS since it was conceptualised. So, I've been with them for quite some time! I look after the data analysts within the team, helping them solve complex business problems and policy challenges.
While a big part of what I do is to help align data with business problems, I also help with the storytelling side of things to draw more meaningful insights for our partner agencies.
I also support a PaaS (Platform as a Service) offering that the DAC has for other government agencies that use our analytics platform for their business analysis needs. This requires me and my team to conduct sessions across the departments to keep the analysts and scientists apprised of the latest that we have to offer in the space.
What type of education, skills and experience are needed to become a data scientist?
I think the biggest fundamentals of becoming a data scientist is that you need to have a feel for numbers and have a flair for storytelling.
I often see people from all sorts of backgrounds – computer science, mathematical, financial, or statistical, transitioning into data science. But then there are a few in our team who come from a media and comms background and really liked playing with data, so they did pursue a master's degree in Data Science to become data scientists.
So really, there's no specific background needed towards becoming a data scientist. There are various academic channels that can be pursued to become a data scientist and then the nuances of it can be learned while on the job as with any other field.
Data science sounds like a very varied field. What are some of the many ways data science can be used?
We see data science being used across a lot of business domains. We have applied it across use cases that are around fraud analytics, risk profiling, or more recently in the areas of media campaign analysis, where you're trying to analyse a certain outcome from a survey that you put forward.
It’s important to note that the business problem is key to any kind of data science solution that we want to put forward. Because you need to have a purpose and a well-defined problem for us to then go and look for the right data and the right solution to align it. So, it can be applied to any domain, what we just need is a well-defined business problem.
To summarize, we can apply data science solutions to any business domain if we have a well-defined problem and good enough data to support it.
What can be done to encourage more women to pursue data science?
Data Science is a relatively new industry, and because of that, it has an air of mystery around it, where people feel that it's very complex. Many people might assume that it's not something that will come easily to them, or that the pathways are really challenging. But that's not what it is. I think the industry needs to demystify it, especially for women as there is a persistent absence of women in the field.
Creating a data science-specific network for women will help with the support and the nudge they might need to enter and thrive in the space. The only other thing to be mindful of here is the velocity at which this industry is changing and hence keeping yourself relevant is a conscious effort that you will have to undertake always.
What are some key projects you and your team have been working on to help improve outcomes for NSW citizens?
One project which is taking most of our time currently is the data response to COVID-19. The Data Analytics Centre works to get data from different sources – either government agencies or the private sector to help with the insights around the uptake of vaccination or booster doses, mobility within the state, spend trends for NSW to name a few.
Another important project the team is currently working on is using natural language processing techniques to evaluate the various government programs. There are multiple feedback channels that the government has for various programs such as Dine and Discover and the Fuel Check app. We are trying to analyse all the feedback that we have received through these channels and combine them together to draw meaningful suggestions in terms of how we could improve our services to citizens.
We recently had a very good session for the fuel check app, and the team have started working on some of the feedback that the citizens have provided, and we are going to continue doing this for the different areas within DCS.
What are the next steps for data science in NSW government?
I think COVID-19 has helped all the different government agencies to come together and build a data sharing environment and understand that data sharing is good and can be beneficial in terms of the insights and the support that you can get from data. But I think this must be encouraged to continue well into the future, and not restricted to the pandemic. From a future perspective, the whole culture of sharing data needs to continue. And people need to understand and value the fact that data can help them with their business problems.
Our small data consulting group within the government also needs to make sure that we try and implement data science solutions that are well aligned to business problems, so people are using data science solutions as a part of their business process and move beyond a proof-of-concept implementation of it.