 Thank you, good morning. Okay, first of all, I'm not gonna use any slides because I just want to talk a little bit about three things today. First is what is government got to do with digital and artificial intelligence? Two is the importance that the ecosystem plays in the government technology as you move forward. And third thing is talk a little bit about your intelligent assistance and artificial intelligence. Okay, just to set the record straight, I've not been a career civil servant. In fact, I've only joined the public sector about three years ago, two and a half years ago. I spent the last 20 years in the private sector so government is very new to me. I came because I was asked to make some changes. When I first joined about two and a half years ago, the first thing I realized was that after about 15 years of outsourcing, the government has zero internal capabilities. It was a scary thing, right? When I walked in there, I realized that we don't build anything. Everybody calls themselves a technologist but actually don't know much about technology because in most of the time when you outsource everything, you outsource capabilities. And we thought that was not possible because if you really want to move into the digital age and to become more relevant, we need to build internal capabilities. And so that was the first thing we did. I walk around the organization. I managed to find about seven people that were doing a lot of scum works. People who were doing their own development, doing small projects, gathered them together. I had a chat with them and they were really afraid because they were thinking that I'm going to shut them down. But instead I told them, you know something, you guys will form the colonel of government digital services. And from that point onwards, we built this organization called Hive at the St. Croller, the one that the Lucas building in One North. We got a space there. I took the seven guys. We built the team to about 40 people and started delivering projects, right? From very simple things like building apps to now building real agile systems for the government. Today we got about 150 people there, all the way from agile practitioners, data scientists, UX, apps infrastructure people. So all of a sudden, we got from literally zero to where we are today, which is building very core digital capabilities within government. And you know, funny thing was that most people thought it was not possible in the beginning, but that's happening. So that's a good thing. The second thing we realized is that, you know, the ecosystem and the community is very important. So one thing we realized very quickly is that most of the time, the way government think of solving problems is that we will build a platform, we will build a system and then life goes on. So when we started creating the ecosystem, we created based on three dimensions. The first dimension is how do you get citizens to participate to solve problems for Singapore? And the first project we did on that, on this concept of what we call digital kampong, I mean, okay, for many of you who are not Singaporeans, kampong is a community, right? You know, the funny thing about the modern living that we have today is that 30, 40 years ago, communities were very important. We don't live in high-rise housings, we live in little communities and everybody helps everybody else. But the funny thing is that as you become more developed, as you build nice condominiums, et cetera, nobody knows anybody. You might be living closer together, density has gone up, but the reality is that actually nobody cares, right? You walk past probably a few thousand people every day and nobody really cares about whether you're alive or dead. So we said, how do we bring people together to work together to solve social problems, simple things for example. So one day I was having coffee with a bunch of people from the Singapore civil defence force. These are the guys that you call when you have an emergency. You call 9.95, they dispatch ambulance to help you. So they were lamenting that actually it's very hard to get an ambulance to a patient in time. Why? Because their service level agreement is about 10 minutes. That means if the ambulance arrives more than 10 minutes, chances of the person dying is actually very high. Because if you have a cardiac arrest and you don't have the attention within a short period of time, chances that you will not survive. So they started adding more ambulances and they started distributing the ambulances across the island to give you a so-called a faster response time. But the funny thing is that despite everything, they still could not meet the service level. Why? Because of traffic, right? Getting somewhere in 10 minutes is not easy. So you know two of our guys and myself were having coffee and we said you know something they might be a better way to do this. Buying more ambulances and getting more paramedics out there to solve your problems. So we built a very simple app and this is where you know three young guys in their 20s sitting down there building an app solve a very important problem. The app that they built was called My Responder. And why is it My Responder? Because it's a crowdsourcing app for lifesavers. So anyone who has medical training, whether it's Red Cross, whether it's you're a doctor, you're a nurse, you're a paramedic, anyone who can do response can enroll as a volunteer in this app. And if an emergency situation happened within a 400 meter radius, you will get a notification on your phone and you can choose to respond. Now it's a very simple idea. The app was built in about four months, didn't cost us much money to build it. And all of a sudden today as we speak right now, I think we've got 14,000 volunteers after about only about less than two years. And what happens is that these volunteers have now saved many lives. Right? Very simple. Something happens, a trigger, a notification comes to your phone, you go there, you do a CPR, you save a person. So this is a very different way of looking at government. It's not all about government producing more, providing more ambulances. It's really about our ability to get people together, build a community. And this concept was initially debunked. Everybody told us it's not going to work because nobody's going to volunteer. But today that's not true because we've got about 13,000 to 14,000 volunteers today. So it can be done. And the power of a digital community is a lot more than most people could anticipate, actually. So this is about building one part of the digital community. The second part we did was also to build an ecosystem of people like yourself. In the past, the government always spent time buying only from the very big companies, the IBMs and the HPs of the world. And we realized that actually this may not be the best thing forward. So we created this open concept called Enolip. What it does is that it's a very simple idea again. It's a matchmaking service. Every month, we get the agencies to come up with their problems. Very simple, I mean, from simple to complex problems, we come together with a bunch of startups and SMEs. Everybody has five minutes to present a problem. So for example, one session could be on the digital assistance and artificial intelligence, for example. So we get like 20 agencies, they come up and say, these are my use cases. Everybody got five minutes to present their problems. Then the participants from the startups and SMEs will spend also five minutes presenting their solution. After that's over, we have a coffee break. They're going to look for each other. They have a speed dating service. You find the person that's most attractive to you in a sense. So you'll find that it's actually very interesting because of this matching process, we were actually able to do a lot of QoC pilots and actually went live with many projects. These projects are not huge projects, but you know something, it allows the community to come, participate and help solve problems. And we also in that process help grow the smaller companies instead of just focusing on the big, large multinationals that we are so used to. So this is the second part about building the ecosystem. The third part goes very much into what we are talking about here, open source. Last year, the digital services team in Hive started this portal called GavPy. GavPy is a concept that we've been working with the US for a little while. You know, there's this organization called 18F in the US. And we work together, we said, sometimes we don't have to build all the applications ourselves. We can actually auction out parts of the application, whether it's a micro service, whether it's an API, whether it's just, you know, a small batch of codes basically. And people like yourself can do a reverse auction, come into it, do the code and we'll buy it from you. And we don't have to go through a horrible procurement process in government because these are all small purchases basically. But what it does is that we are now tapping on a huge open source community to help us build the codes for government. So it doesn't have to be something that we do ourselves all the time. And this has worked well. We want to push it a lot further. We want to make the limit higher because I think today the limit is about five grand, five thousand. Our intention is to push it to up to $20,000 actually for each person to come in and do the reverse auction. So again, the whole idea that we are pushing forward is about building communities. It's not about us doing everything ourselves. And I think this open way of doing things has changed the way the government reacted to many things. Because in the past, everything was done internally or is outsourced to a massive vendor. So last two years, things have changed a lot. So I see a humongous future for the open source community. I see a lot of opportunities in terms of how individuals or small groups of people can contribute to the government. The last thing I want to talk about is this topic on machine learning and personal assistance. We started this about two and a half years ago again when I first came on. We realised that actually a lot of the work of government is answering questions on policies and on, you know, basically FAQ stuff, you know, you call a call centre up, somebody's going to answer the question. And we find it to be very, very ineffective. And it's not good use of human resources. So we started looking to create an intelligent personal assistant, a digital assistant. And like everything else, we went to two ends, right? The high end way of doing it was that a bunch of people in government was very enamoured with Watson from IBM. So they brought in IBM, brought Watson in, and Watson wanted a whole bunch of money to even do a proof of concept. On the other extreme end, we got a bunch of people like you who said, nah, that's bullshit. Let's try something different. So we spoke to a bunch of start-ups, and we got a small company to come in. I'm not going to mention the name, but it's a very, very small company. They came in, and they did a proof of concept with us, and they implemented in two agencies in three months. And each of the implementation causes $20,000. So that's basically what it's all about, right? Watson wanted to charge us $5 million to do that. And after one year of Watson, we had done nothing. But these companies, one small company came in, two agencies in three months each $20,000. Now, I think that's what's all about today, right? It's not about getting to a big company and life is good. That's not about that. It's about getting the best solution for the purpose. We started very simply with an intelligent FAQ system, as I said, that cost $20,000. Today, it's natural language processing. Today, we're injecting more intelligence into the platform. We're working with Microsoft. We're working with AWS on Lex. We're working with Google on TensorFlow, etc. So that's how we build. We start something small. We incubate that thing. We add intelligence to it. And most of these things that we're doing are open source. I mean, Amazon came to us and say, hey, try the Lex API. Let's make it happen together. And sure, we are not partial to any particular vendor. In fact, if it's open source, it's even easier for us because it becomes easier for us to develop over time. So the AI journey or the personal assistant journey for government has started that way. We have a really small company. The only problem that small company is that after they got successful, we had a problem. We got 35 agencies clamoring and said, I want this too. And the owner of this company came to me and said, you know something, I got a problem. I said, why? I've only got 12 staff. How can I do all these things? So this is a challenge, which I think as a government, we're also keen to help out. How do we make this small company scale up and be successful? And I think the future is in this area, right? It's not just about buying from a big company, but getting a small company, giving them the opportunity, making them successful, and then helping them scale up. And I think, you know, with that, you'll find that the open source community has a lot of work and has a humongous contribution to the way the government can work in a future. So with that, thank you very much. I hope you enjoy the rest of the day. Thank you.