Canadian Government Executive - Volume 24 - Issue 01

January/February 2018 // Canadian Government Executive / 29 THE INTERVIEW I could tell you if it was Oracle, if it was Microsoft, if it was IBM, and which one was Open Text. And I could do that in Ot- tawa as well. We all know it, right? So as opposed to declaring a technol- ogy as standard and building around and declaring it, saying, “Okay, well, we have 300,000 users of this thing. We will not rip and replace this in the next ten years.” In- stead of defining our solutions when go- ing out for procurement, we may want to flip the scenario and outline the problem and have suppliers propose viable solu- tions. By putting the problem out there – be- cause we know our problem – and by opening it up, you get answers. I don’t care if it’s an academic institution that an- swers my problem in scales. That would be quite a concept, because then you’re creating value through procurement in digital and innovation in the country. To say that we’re fully transparent – sure, yes, we are. But if you look at the require- ment, and it says, “must integrate with this, must integrate with that, must integrate with this,” that’s not transparent. You’re after a specific thing, and we’re using the word ‘transparency’. But by say- ing, “Hey, I’m willing to communicate, and it has to meet certain new standards, and potentially be open source” – well, then you can actually pivot. You could be a bit more nimble and a bit more agile. Not in a development sense, but in the ac- tual physical sense. And you can look at procurement dif- ferently. Again, we did it in two months by putting the problem out there. We got four completely different solutions. It wasn’t four of the same. And then you’re looking at price. We had four different solutions: One of them was using AI, and another one was using another thing, and so on. Four different solutions. How amazing is that for value? Q: What was the project? It was launching a pilot for ‘open by de- fault’ with four departments: Canadian Heritage, Environment and Climate Change Canada, Natural Resources Can- ada, and Treasury Board of Canada Sec- retariat. We’re starting in small steps. “Re- lentless incrementalism” is a line that I’ve heard from one of my mentors that we’re going to use all the time now. Innovation does not discriminate. The excuse from government of, “Well, they’re not big enough to scale.” Well, maybe we’re too big to be agile. Why can’t you flip it? I think Airbnb started with roughly a dozen people, and it completely disrupted the entire global accommoda- tions and hotel sector with twelve people. So don’t talk about scale, because now the value in scale is being small and nimble. It’s not about being large and doing these 20-year projects. You want to look at gam- ing with Angry Birds. Two or three peo- ple, right? Minecraft. Same thing. A bil- lion dollar acquisition by Microsoft, and it’s two or three people in a basement. It’s not about size anymore. It’s about impact. Q: Fascinating. Let’s back up a little bit here. Tell us about your mandate and your areas of focus? Right now for us, it’s just about return- ing to basic principles of good technol- ogy management. We had stopped doing enterprise architecture planning, so we ended up with hundreds of versions of the same system or the same function. We have close to 400 case management sys- tems in the federal government. So if you look at 43 departments, that means that each department has nine case manage- ment things. How are you supposed to manage systems? How are you supposed to be agile and nimble when you don’t even have the same base or the same platform? The an- swer is: you can’t be. So we’ve got to get back to basics, and that means saying no to things and consolidating things. And it’s painful. We have to take the pain and, because we’ve created a situation for our- selves where we had 8,700 applications across the Government of Canada. That’s over 200 applications per department – big or small. So basic housekeeping is one of the things that we’re looking at right now. We have to get back to good management technology and making decisions based on business means, not based on what our infrastructure can support. So start with business, look at the information and the applications, and then the technology. The other part of this is that we have a lot of lessons-learned to look at from some of our IT projects and our failures at ex- ecution. That’s also good housekeeping around project management. We, in a lot of cases, don’t have the right ownership or governance models for projects from the get-go. So how do we change the focus at the beginning of the projects to ensure we have a business owner and a project team answering to the business owner and working together? That kind of ba- sic principle and practice has gradually disappeared in the last 10 to 15 years in tech in Ottawa right now. Again, it’s part of good housekeeping. Some people love that – I do, because I’m a military guy, so I crave structure – but then there’s also the threat, right? AI job displacement, 360,000 hours of ac- counting work done in three seconds at JP Morgan last year. It’s not an “if”; it’s a “now”. So what happens when the banks all deploy AI in Canada over the next two years – which they will do. How do you regulate something where you can’t even understand it or talk to it as a sector? And that’s not just in the financial sector; you can pick any other sector that’s moving quickly on blockchain or AI or any other kind of exponential technology – we need to get a grip on that. The second priority is actually leapfrog- ging and not trying to do things in a linear way, but actually getting to some of this exponential tech in a quick way. The fed- eral government just invested hundreds of millions of dollars in AI across Canada, and I haven’t even touched on operation- alizing some of that investment. And then another aspect is just working differently and challenging the culture. So that’s the third priority, and that goes to an agile procurement process. It goes to doing things with partners more efficiently – all of that stuff. Q: With our current model of public service – the Westminster model – our organizations tend to be quite hierarchical and siloed. This departmental approach to