 What I want to talk about today actually is a bit different than what the previous speakers have said because indeed my background is in software engineering, but that was back in the days when it was easy to program, quote unquote, and we wrote simple code line by line. And, you know, as a consultant, we just sold the hours. Nowadays, there's a lot of interesting possibilities as you've heard from the previous speaker when it comes to data and AI. I've actually always been fascinating about it, fascinated by the difference between the potential of technology and how it's actually being used and how we leverage that potential. So that's what I want to talk with you today. So I won't be talking about AI technical aspects on how to do it. Rather, I would show you a challenge and I would ask people here in this group to think about how you could help address that challenge. Very quick, what I see as, you know, one of the core things that we do in information management is in a crisis situation. It's fairly easy. We try to make sure that we have the right people, giving the right information, the right time with the right in the right form that they can use it. That sounds pretty straightforward and it is, but when you think about crisis situations it quickly becomes complex. In the initial stages, when something happens, there is this high level of uncertainty and very limited information that we have available. And as the crisis goes on, the uncertainty drops and we have more and more information means we can make better choices. So my favorite example, walking into a theme park in the Efteling, you've never been there before. There's, you slowly start uncovering the park and its layout, you know what rides are cool and you will learn to, you know, what is worth the queue for or not. Although I would not recommend going to the Efteling, especially on a hot day today, and with the crowds. But as the crisis evolves, things become more stable. So it means you can make better decisions. The goal is to reduce this uncertainty and increase the relevant information for a decision maker. But the problem is that when you talk about disasters, there is this situation that we haven't predicted. Yes, people have talked about pandemics people have talked about crisis that could happen, but it's how exactly it unfolds because the exact effects are we don't know ahead of time. That's why strict scripts of dry book in Dutch don't work in a crisis because it's never exactly what you prepared for. So it also means there's no time you can't sit and say okay well let's start developing an algorithm for this or let's, you know, take our time figuring out what it is, somehow there is a pressure required for you to act. There is no plan and I know if people are in crisis response they would argue no we have plans, especially in the Netherlands we have too many plans. But there is not the right plan we need we have building blocks we have capacities we have resources but how we use them. When do we deploy them. It's not fixed and you've all experienced this in the last one and a half year, how that works and what the challenges are. It's not so clean it doesn't look like information the uncertainty goes down as the information goes up. It fluctuates over time, we get more information we have higher uncertainty. And in fact today, we have so much information that people can pick and choose what they want to hear and what information to use. It's this information overload and the complexity of conflicting information. What is the challenge. And what I also think I would encourage you to think about is, when we talk about this potential of AI we quickly look at the data we look at the technology what we can do. But there are the other aspects that make it work or not, which are the people and the organizations they operate in, and with organizations I also mean communities, any kind of network of people. So what I want to do is quickly take these apart using the COVID-19 example here in the Netherlands because last March, I joined the what's called the National Operational Team Corona LOTC, together with some students as an information manager in the early stages of this COVID response. And I want to show you a bit of the reality of what happens here. So, first, a little bit about that data. If you imagine in the early stages of a crisis, what is the kind of data you get. Well, it doesn't look like spreadsheets or data sets, it looks like these kinds of messages. So, this, these are people just calling us emails, things being mentioned in a meeting. It is not structured data, it's signals, it's small qualitative pieces of information that are coming in. So what do we do with these things well we try to come up with some classification for them. Is this a problem, should we keep an eye on it, our problems coming. For example, you will might remember that last year around this time we had the cell towers being lit on fire because somebody was spreading news that 5G was causing Corona, or something along those lines. So all of a sudden the fire department also had to deal with that while they're being affected as well. It's not something we can find in the data but something we have to monitor. So what we did is we came up with the classification saying okay well let's break it down into simple things. So we say, you know we use some color schemes, is this controllable, is it something we should monitor, is it, well, part of my English completely effed situation and should we act. And is this something stable, is it improving, is it getting worse. We made a simple classification and that allowed us to make these kind of simple overviews and this I'm talking here about less than a couple of days to make this. And what you see here is not a system is not something fancy. This is PowerPoint in the background so these pictures are made in PowerPoint. I can tell you they ended up at pretty high political levels, based on those small messages. We made these overviews, and people relied on that. Does that mean there is no data at all in crisis duration know that's not true we have data, there's the humanitarian data exchange that has a lot of data sets that we can build on. But the next challenge is the technologies we use. Can we work with all of them again in the first days we're working with pieces of paper, and things like Trello to get things organized because it's easy if you get people from different organizations together to work on it. In the first days in a crisis, it looks like this posted notes on a flip chart. You start classifying building basic classifications. So this is more on the structuring and organizing it in a way that everybody in the organization can work with it. We need to have everybody use the same language and the same categories the attached to information, then we trying to build clever systems where we get more work. In the end, of course we start thinking about okay once we've established this how can we streamline this how can we automate it. So we came up with a basic enterprise architecture if you want to call it, essentially collecting the input which could be structured possible data could be phone calls could be meeting notes from different partners coming in. And then, you know, start building that initially in Trello and you will see slowly we start replacing things with more technical tools, like modeling system dynamics modeling agent based modeling and all these kind of tools to help us get more understanding but here we're already talking two months into the crisis. This is against a key point is that we start realizing that we don't have to do all these things ourselves and I'll come back to that in in at the end as well. We had a colleague of mine was working at Delft University as a data manager or data stewardship coordinator actually. So we had all these people contributing data and using it from us, and we said how do we handle it so we asked them, you know how to do this can you write us a policy document on how we should do it. So it is much more, especially if you work for the government in crisis situation you have to think about how to handle and do this safely because you're under the public eye. And I guarantee you sooner or later there's going to be an audit, and people can ask questions how did you reach that decision based on what and you have to justify everything you've done. As you probably seen in the news sooner or later the government will find out what you've done with the money and ask how what's going to happen with it. So the key point here is essentially what we've learned that works is the people, because we can bring in all kinds of tools, data modeling, but it only works if we have the right people in the organization. And the majority of people in a crisis team nowadays are people who have not been trained as data managers who have not been trained with AI or data science as a background. There are people who are operational firefighters people who handle crisis. But only have they don't know what is possible but they can also not value the potential or even assess how it should fit into the operation. These are five students from Tilbury University that I invited into the national team Corona to do all tasks around information management data management modeling tools technical administrative support. There is no flexibility to to adapt to the situation. Oh, here's a new tool we don't know how it works, we'll figure it out. There is a different generation currently employed which who is more like being trained, you know, you don't figure it out you being told how to do it exactly, which removes a lot of the flexibility. And especially bringing in the younger generation is I think a key aspect, because you can bring in all kinds of smart tools, but people are not able to assess the value of it, or might be even relying might not even understand, if you say to some deliver results with them saying well we there's a 50% confidence or 50% validity to this, what that means. So it is really the people that are the difference between having something that has potential, and actually making a difference, hopefully for the better, but at least not for the worse. So, maybe this is, I think where Tilbury University, and this group especially brings in a lot of interesting perspectives, because the opportunities are there when we talk about data and technology, those are being developed and we've seen great challenges, but the next challenges is how do we leverage that into an organization how do we make sure that if we're dealing with this high uncertainty, and this unexpected situations that we can leverage that potential. So, in the end, I will skip that in the end I think the question is, you know, we have this wonderful tools and hammers. How do we put this into a toolbox so that people can grab it and use it, and not just blindly following the one tool but understand there's a whole range of options that they can deploy. And more importantly, how do we turn people into effective builders with these tools and crisis situations. I thank you for your attention.