EXCLUSIVE – Staying on the Edge: Origins of the NSW Data Analytics Centre
Could you tell us about your role at the NSW Data Analytics Centre?
I’m the Director of the NSW Data Analytics Centre and I’ve been involved in it before it existed. I was involved in the drafting of the legislation, the business case to create it as well as the day to day operations. I report to the NSW Chief Data Scientist and CEO of the DAC, Dr Ian Oppermann.
What are some of the major initiatives that are happening at your department?
We have ongoing projects endorsed by the NSW Government, which are complex data problems. They involve data from multiple agencies with complexities that a single agency couldn’t solve by itself. For example, if it’s a transport problem that could be solved with transport data, Transport for NSW have a great analytics team that can solve it themselves. But if the transport problem requires education or health data to solve that problem, then we can help them. And in all projects like that, the agency that’s actually responsible for doing the implementation, are the heroes of the day. They are the ones who are going to make changes in the work that they do based on the work that we do with them. Our projects range from things like determining whether a building fire alarm is a false alarm or a true alarm – whether it’s burnt toast or a real fire, down to the other end of the spectrum where we look at issues around child protection and out of home care. So we cover a full spectrum of projects. Currently we have 22 live projects.
When was the NSW Data Analytics Data Centre first started, how many projects did you all begin with?
The NSW Data Analytics Centre was first announced in August 2015, so we’re just over a year old. The first thing we did in terms of projects was that we put up 9 projects to Government for endorsement. We had expected them to approve maybe 3 or 4 of the projects and they came back with 10.
How big is your department/team currently?
We’ve got less than 20 full-time staff and we also have a number of students. We are training the next generation of data scientists as well. We have an arrangement with a number of universities in NSW through which we bring in their data science students to give them experience, and that helps us add firepower as well.
In terms of the data your department is handling from multiple agencies, how do you manage security and data privacy from your end?
We comply with privacy and security the same way any other NSW Government agency does. Privacy and confidentiality are at the core of our operation, because if we make a mistake with privacy, we would lose the trust of the agencies and of the public, and we don’t want that.
The NSW Privacy Commissioner was a member of the steering committee that guided our establishment, and we continue to work with her office on individual projects as different projects require different data sets. We have also met with the Federal Privacy Commissioner to ensure our work fits within the legislation.
Your department is very young and new, being just slightly a year old. What are some of the challenges involved, especially when the department is still at its infancy stage?
Data sharing is very much a challenge of cultural change. One of the big challenges is the silo mentality within government agencies and people who want to hold on to the data. In some cases, they’re reluctant to share the data because there’s personal information in there and they need to protect it.
In NSW, we have legislation to remove barriers that impede the sharing of government data with the DAC or between other government sector agencies. That certainly helps start the conversation about why agencies should share their data with us.
So basically it’s difficult to get agencies to willingly share their data?
The culture, at the highest levels, has changed and some agencies are more willing to share their data, but there are still some pockets where there’s a bit of resistance. We would expect that with a change like this.
So apart from the siloed mentality, what other challenges do you experience?
There are some agencies that don’t fully understand their own data. Although they would like to help us, they don’t know what they have or they may not be able to get access to their data. They may know that the data might be relevant but they are not sure either how to get it out or how to bring it out in a useful format.
We often sit down with agencies to help them understand their data or help them work with the data. That helps them see why the data is important or helps them give it to us in a suitable format.
It’s interesting, I guess part of your department’s role now is to educate the agencies to work with their own data?
Our purpose is two-fold. One, doing the priority projects that have been approved by Government, and the other is to raise the level of analytics across the government.
Although we are specialists in government data analytics, we don’t want to be too large – 20 - 30 people is the right number for us so we can be agile enough for the projects we need to run. We don’t want to be a 100 strong. That will make us unwieldy. We want to stay small and raise the level of analytics of other government agencies.
We’re doing good analytics, we want to bring them all up so they’re all at the highest level. In some cases, the agencies can then start running more complex projects and do more complex analysis themselves because they have learnt from working with us. In some cases, it’s just helping them understand their own business so they can go on and fly by themselves. We also learn from the agencies that we work with and together we will all get better.
What are your thoughts on innovation at your department?
The Data Analytics Centre is effectively a start-up within government. Very few people in our team have worked in government before. Most people in the team are from the private sector or academia; they bring a completely different view on service delivery within government, and so just by our nature, we’re being innovative.
Since the NSW Data Analytics Centre is quite new, do you all also collaborate or work with data analytics centres of other governments in Australia?
We recently had representatives from ACT Government visit us. They wanted to talk about how they can undertake the same journey in government data analytics, and of course being our closest neighbour, we want to help, we want to share what we do with them and how we are doing it so they don’t have to make the same mistakes. They can just pick up and start from where we are. Western Australia, Northern Territory, South Australia, Victoria and Queensland have also come to speak to us.
Internationally, we have engaged with some of the provinces in Canada; multiple delegations from Thailand have come to visit us; we’ve hosted delegations from South Korea and the Singapore Government. They are interested in our legislation, they are asking questions about how it works, the processes we have gone through and how we approach our problems.
Is the NSW government the first to have a Data Analytics Centre in Australia?
We are definitely the first dedicated whole-of-government data analytics unit in Australia. We are the first in the world to have the data sharing legislation to support the work we do. Other departments and agencies internationally have similar analytics capabilities but none of them have the legislation to back them up. We also have an advisory board made up of largely non-government people, and they keep us “honest”. By that we mean they encourage us to stay on the edge, we want to be ahead of the curve. On our advisory board, we have people like the CIO of the ASX, senior people from universities, people from Google, people with a lot of experience working with data in their own companies. We want to learn from them and work with them to stay ahead.
Since your department is quite new, where does the funding come from?
In our first year we had some seed funding from within the Department of Finance, Services and Innovation to get us up and running. In the 2016 NSW Budget, the government announced 4 years of funding. It gets progressively smaller each year so we have to start becoming self-funding as we grow.