Fraser Charles, Partner Ecosystem Leader at Dataminr, explains how AI and location intelligence work together to manage risk.
If you liked this episode, please share it with a colleague.
November 18, 2026
NARRATOR: Welcome to the Esri and The Science of Where podcast. Managing risk depends on understanding relevant information. Weather patterns are vital for wildfires, but weather doesn't matter to a fight breaking out in a stadium. Knowing what matters is how Dataminr assesses risk. Dataminr uses an AI platform to analyze billions of pieces of public data from global to hyperlocal sources.
That's how Dataminr's customers manage operational, cyber, business, and reputational risks. And Fraser Charles, partner ecosystem leader at Dataminr, says that the power of where is central to understanding those risks.
FRASER CHARLES: So much of what we support with our mission ties to physical locations, and it's just such an intrinsic part of understanding any situation as it develops, where almost all of them are totally dependent on the critical context that location awareness actually provides.
NARRATOR: Fraser Charles explains why half of Fortune 100 companies use Dataminr and Esri software together to deliver analysis in real time.
JOHN LENAHAN: Hi, Fraser, and welcome to Esri and The Science of Where podcast.
FRASER CHARLES: Hey, John. Thanks for having me.
JOHN LENAHAN: So Fraser, you're the director of partner ecosystems at Dataminr. What does that mean?
FRASER CHARLES: I oversee a pretty broad portfolio of partners. In this particular portfolio that I look after, we consider a technology partner, and for us, those are software companies. Those software companies are prioritized for our ecosystem based on the ability for us to create an integration with them, like Esri, and that integration leaves a customer further ahead than if they were using our products separately.
JOHN LENAHAN: So you've got a product, in this case data models, and we'll kind of dig into that. And the software companies are really a vessel to serve that up to a customer.
FRASER CHARLES: A vessel's a perfect way of saying it. When a customer is trying to get the outcome they are, their stack needs to be highly interoperable across all layers, and so that's really the goal for us.
JOHN LENAHAN: Okay. So first, let's understand very simply how Dataminr works.
FRASER CHARLES: Dataminr is real-time event, threat, and risk intelligence. That's our sort of modus operandi. We want to deliver the earliest actionable intelligence on breaking events, emerging threats, and unexpected risks for our customers.
If you think about what we're actually trying to do, we're trying to characterize the global threat and risk landscape for them. In order to do that, you have to capture signals. You have to capture those risk signals, and then you have to evaluate what they mean, because that's what the customer needs to know.
JOHN LENAHAN: So let's talk about that in the context of a user. A firefighter or a risk analyst for a company. What do they get from this?
FRASER CHARLES: If you think about what types of organizations need our help today, there's defense and national security, there's critical infrastructure like airports and ports, state government organizations, as well as enterprise. If you think of an organization, they've got physical locations, there are people on premises at those locations. Companies that operate a transit system have an obligation to keep both the operators and the passengers safe.
From an emergency management point of view, let's use your firefighter example. They're trying to respond to some of these challenging crises and situations.
JOHN LENAHAN: But there are other cases where your information and the threat intel you're providing is valuable and useful when it's not such a significant event. Can you talk about some of those examples?
FRASER CHARLES: Right. When you think about the scale of the problem that we're solving, we are typically most relevant for these global organizations. And when you think about the remit that they have and the breadth of what they're trying to oversee across country boundaries, in different cultural environments, things like that, the frequency of high-impact events is higher than you might think.
However, when you think about more local events or risks that we would help with, I'll give you an example. A small fire on a corner, there could be a chemical spill, and it doesn't have to be of huge scale, like an industrial spill that shuts down a port. It could be something like a traffic accident, and one of the vehicles involved is a truck carrying chemicals. Those are some examples.
JOHN LENAHAN: Going back to what you said in terms of the global scale and how that's a real target for your products, it makes me think of something your colleague said, that the flood of data is kind of buried in a digital haystack. Can you, in your own words, explain what that means? What's the opportunity?
FRASER CHARLES: Again, going back to what we talked about a little bit earlier, trying to characterize the global threat landscape, we do that today by ingesting over one and a half million different data sources. Those can be social media sources, information coming from the deep and dark web, news sources, blogs, IoT sensors, connected devices. So that in and of itself represents a haystack.
JOHN LENAHAN: Absolutely. Overwhelming.
FRASER CHARLES: Overwhelming. And so what we do is understand, at the level of a customer, what is it that you're trying to solve for? Because you just mentioned a relevant point. Some folks are more focused on operational continuity, business continuity. Some folks are focused on public safety and security. Some folks are focused on traveler security. What that customer cares about is going to be a function of what's their job to be done.
What we do is basically allow dials and equalizers, if you think of a music player, and they can tune what comes out of the machine. So what's going in? One and a half million unstructured data sources: sound, video, text, sensor device signals. And what comes out the other end, we're able to highly tune to what a customer cares about.
JOHN LENAHAN: And that's a perfect segue for the role of location technology. Because part of that tuning is also where, right? Can you expand on that?
FRASER CHARLES: Absolutely. I would say it this way, John: Location is one of, if not the most critical piece of context for a customer. It's one of the key filters that we apply to what comes out of the machine. If they don't understand where this event is happening, where this threat is emerging, how do you make a decision? How do you deploy resources? How do you respond to a crisis? You can't. So it's super critical.
JOHN LENAHAN: Coming back to the point you made earlier about traveler safety, a traveler's going from point A to point B to point C, and they're traveling somewhere. If something's happening along that travel route, that's a risk that's important to them.
FRASER CHARLES: One hundred percent. There's a bus traveling through Manhattan, and they are going to care if they're within some defined proximity of this risk event. With a combined product of ArcGIS Velocity and Dataminr, it's very easy to achieve a scenario where you could reroute that bus based on a severe risk event in close proximity. Now you keep people safe, the operators safe, the passengers, and you don't slow down your transit on-time performance.
JOHN LENAHAN: Another example of that would be high-value routing of a shipment.
FRASER CHARLES: One hundred percent. Another example is critical deliveries of sensitive materials, things like that. We actually support folks doing that today.
JOHN LENAHAN: So you talked a lot about emergency, you talked about some transit. I'm also interested in the concept of employee safety. Think of a big city, someone's in the lobby with erratic, kind of dangerous behavior.
FRASER CHARLES: July 2025, there was an active assailant, and we generated the first alert of that risk within six minutes of the situation starting. The feedback we got from our customer about that specific event was that they were in an adjacent building, very close in proximity to the event happening. So close that they had people down on the street engaging with law enforcement officers, and they were receiving better information from the Dataminr alerts coming to them describing the scenario.
JOHN LENAHAN: In that case, it's both safety to people that are in the building, but then it's also, like you said earlier, the responders.
FRASER CHARLES: Absolutely.
JOHN LENAHAN: Logistics firms, cargo ships, trucks, delivery trucks need to know the weather systems that they're traveling into and through. But can you tell me a little bit about less obvious forms of transportation, like public transit, using the data you collect? You gave the example earlier about New York City, but are there other examples as well?
FRASER CHARLES: That's an interesting point. The earlier examples that you just mentioned, the logistics firms, the ships, I think that's one of the interesting things for me about the nature of the data landscape that we are using to describe the threat landscape. There are maritime sources. There are FlightAware tracker alerts that we're consuming as well in order to describe events that are relevant for those folks that you just mentioned.
From the other end of the spectrum, say the transit mode of moving people around, they're caring about the safety of passengers and operators. They're caring about on-time performance. They're also caring about emergency preparedness.
One story that I've come across is there was a significant traffic viaduct that was shut down for maintenance, and they had to reroute traffic around that, make a detour through an area of the metropolitan area that was not designed for that sort of traffic volume. They were now concerned about having an inordinately high volume of traffic going through this area. Again, safety becomes an issue. They're looking for both volume in that sense. They're also looking for significant incidents that they would need to deploy responders to.
JOHN LENAHAN: Dataminr works within an ecosystem of partners and provides, I guess, intelligence and information to other software platforms and other capabilities.
FRASER CHARLES: That's correct.
JOHN LENAHAN: At Esri, we talk about geographic information systems, GIS, being a system of systems. What do you wish people understood about location technology and leveraging data within a location technology system?
FRASER CHARLES: We have a unique perspective because we work at a company that produces this software, so you see it. I've got it on my phone. I've got my alerts going off and I'm looking at those very frequently.
But I think that is one thing when you step back and actually think about the technology that's producing the alerts that are coming through on my phone. It's so powerful just for that visibility and awareness of what could be around the corner that, if you had that visibility, you could easily avoid.
In some sense it sounds rudimentary, but just the awareness that that type of technology exists and it can be so valuable and preventative for negative impact of some of these sorts of emergencies.
I think the second thing is, if you think about the power of real-time intelligence, I would also say the power of location technology. Again, this is what we talked about a little bit earlier. So much of what we support with our mission ties to physical locations, and it's just such an intrinsic part of understanding any situation as it develops, where all of them, or almost all of them, are totally dependent on the critical context that location awareness actually provides.
I would say the second thing, John, is it doesn't have to be a heavy lift to consume it. Dataminr is fundamentally an AI company, and there are different forms of AI that can be consumed by organizations. Some are going to require a much heavier lift. This is something where Esri has out-of-the-box solutions. Dataminr can fit right into those out-of-the-box solutions. You can have an incredibly valuable common operating picture stood up within weeks.
JOHN LENAHAN: So that's speed to value.
FRASER CHARLES: Speed to value. If people understood that the nature of the solution is available, it is not a heavy lift, I think that's what I would leave it at.
JOHN LENAHAN: So you talk about the simplicity. I want to talk about the use of real-time data, location technology, and how vital they are for emergency response, like the example of wildfires. Can you expand on that?
FRASER CHARLES: For sure. A wildfire is an example. Most of these risk events that Dataminr is characterizing for organizations are dynamic by nature. They're constantly changing and evolving. I think wildfires are a perfect example of that.
Folks who are involved in the organization that's responsible for preparedness and planning, maybe they're part of the agencies that are involved in response, maybe they're involved in rescue, all these different things. They have to understand the changing conditions in real time in order to make the right decision.
There are different stages of these events that are important for them. There's: Give me the earliest warning. Help me with decision support. And lastly, help me respond effectively to this type of response.
Take the wildfire example that we're talking about. Variables there: weather conditions, public activity in and around the area, road traffic. If you're going to deploy resources, can you even get them there safely and effectively? Growth and advancement patterns of the perimeter of the fire, obviously.
All of these things are combining, and so the degree to which you can provide higher-fidelity insight in real time as these decisions are having to be made, the benefit of the combined solution is a time advantage and an information advantage. Nobody in a situation where they are responding to a crisis is going to say no to either of those.
JOHN LENAHAN: How I also like to think about it is these responders need a shared understanding. And how do you create that shared understanding? One of the best ways to do that is through mapping and mapping technology. It's just a natural way to get around the table with your digital tools and technology and understand, how do we act on this threat?
FRASER CHARLES: It would be in very rare cases that a single agency is responding to this. You're always going to be coordinating across multiple agencies. And so what is the common operating picture to share the information, collaborate, and then coordinate?
JOHN LENAHAN: And that's an obvious thing when you think of wildfires. But if you think of New York City, a large bank, if there's an incident at the bank, immediately they're going to be coordinating with NYPD, emergency response, and they all have to share and collaborate as well.
FRASER CHARLES: Exactly.
JOHN LENAHAN: So you mentioned AI earlier, Fraser, and the use of AI to sift through all the sensors and the input data. Can you talk a little bit more about that? How does that work? What are some of the ways AI is helping you make sense of it all?
FRASER CHARLES: For sure. We talked a little bit about the breadth of what we're ingesting to describe this risk landscape. Those more than a million different sources are coming from social media, IoT devices, like I said. Fundamentally, it's a massive-scale data problem. If you abstract it to the most fundamental layer, that is a perfect application for AI.
If you think about what we're doing at our core, we're recognizing patterns. We're identifying anomalies. It would be one thing to do it with a human-based approach if you were doing it in a very constrained category of risk, if you're doing it in a constrained region, and the volumes were low.
Because we're serving customers globally, that is fundamentally why, since 2009, we've been working on, I think it's more than 55 different AI models, designed and built by a team of hundreds of AI PhDs and machine learning specialists.
JOHN LENAHAN: And really the volume and the velocity of the data necessitates AI.
FRASER CHARLES: Exactly. It's the scale of the data, and it's the speed at which we need to meet the customer's requests.
JOHN LENAHAN: Thoughts about the future of incident alerts and risk? Where do you see things going? How does the technology change? How are you going to change?
FRASER CHARLES: The enabling technology is changing so quickly. Even when you think about what we're able to do today with agentic AI compared to the real-time alerting, which in and of itself was impressive and very useful to our customers. It's changing so quickly that I think, near term, external-internal data fusion. Longer term, this sort of predictive intelligence mode.
JOHN LENAHAN: And getting to that predictive or the action, could you see a future where your data is being brought in, you're providing that living intelligence and that intelligence brief, but you're also building a system that knows when an event like this has happened for the last three years, what you usually do next is this, let's just actually trigger that as an action?
FRASER CHARLES: Totally. Tying it to standard operating procedures that are well-defined. If you looked at that pattern that you just described, John, and you could say with certainty, 99% of the time we go do this, well, then just trigger it. You can take the human out of the loop.
JOHN LENAHAN: And that's that speed, right? Because the difference between setting up an alert and notifying a regional director of a facility, the difference between five minutes and 20 minutes, or 10 minutes, can be big in a life-and-death situation.
FRASER CHARLES: It can be huge. Absolutely.
JOHN LENAHAN: Well, Fraser, this is wonderful. Great conversation. Thank you so much for joining us, and we'd love to have you back someday.
FRASER CHARLES: Thanks so much, John. Really enjoyed the conversation.
NARRATOR: Thanks for listening to the Esri and The Science of Where podcast. If you liked this episode, please share it with a colleague.
Fraser Charles, Partner Ecosystem Leader at Dataminr, explains how AI and location intelligence work together to manage risk.
If you liked this episode, please share it with a colleague.