Earth Observed | Episode 19 | Boosting SME Logistics Profits with AI

Speaker A: Hello and welcome to the Earth uh, Observed podcast. A podcast where I sit down with the founders, the technical leaders and the minds shaping the future of Earth, uh, data, geospatial technology and atmospheric AI across Europe, North America and beyond. In each episode we take a vertical dive into a topic chosen by our guest, giving you a front row seat to the ideas, the tools and the challenges that enable innovation, from space to subsurface and everything in between. Today I have the pleasure of being joined by Selby Carey, co founder and head of product and engineering at uh, Mapify, a tech startup helping SME logistics and warehousing operators boost their asset profitability through AI. Selby, thank you so much for joining us today. How are you?

Speaker B: Very good, thank you.

Speaker A: Good. Fantastic. Fantastic. So look, um, as uh, the title goes, uh, we're going to be looking into mapping logistics assets, enhancing profit margins for SME logistics operators with AI. Before we dive into the discussion there, Selby, I was just wondering if you can give our audience, the, uh, the listeners, the viewers there, just a bit of a brief teaser into some of the topics that we might uh, cover today. Sure.

Speaker B: So what we'll learn today is the challenges surrounding the logistics industry and the opportunity that AI provides, the complexities of actually deploying AI into an industry that is traditionally pen and paper or Excel based. And also the opportunity in the future for how AI or automated systems can help us run our entire logistics operations semi autonomously.

Speaker A: Amazing. Amazing. I'm really excited about this particular episode. Reason, uh, being, you know, a lot of the time I have people coming on from, you know, a climate risk background. Uh, maybe they've worked in like atmospheric AI to do with the weather and meteorological data. Uh, maybe they do come from a true space background. But I really am, um, interested in what you guys are doing with Mapify, um, in that logistics realm. Um, and I think there could be, ah, a lot of viewers, a lot of listeners interested in the type of work you're doing and how you are answering particular challenges, uh, there. So I guess starting off, uh, from the beginning, a very simple question. What is Mapify?

Speaker B: It's a good question. So you can have a look on mapify logistics.com it is a visual productivity tool that manages your entire asset operation. So what this means in practice is you might have a small logistics business or you know, you run operations that incur some sort of cost. You may use equipment in a warehouse such as a forklift, or you might move goods via, uh, HGVs or large trucks. So it's a range of different operations you might perform. And each operation has a cost, it has a revenue associated with it, and it has potential conflicts or problems that occur along the line. And to try and juggle all of that in your heads or on a spreadsheet can be quite complex. So what Mapify does is it aggregates all that information and it creates an operating system for your organization. So you can see it on a map, on a diagram, you can see a dashboard of how your operations are performing. And you can schedule regular tasks and prevent conflicts before they occur. So what this means is that you can say goodbye to downtime events where your operation stops and you're unable to make any revenue. And you can say hello to profitability, because the profitability is what helps SME to grow. And that's what we focus on, helping small businesses grow. Using AI.

Speaker A: Of course. Of course. It makes complete sense there. Tell us, what was the inspiration, what was the reason, the meaning, the purpose behind, uh, starting Mapify? Because I understand that you also have a co founder as well, is that right?

Speaker B: Yes. So Hamish and I, Hamish might be my co founder and the CEO. We actually met about two or three years ago and at the time I was working on a different venture and, uh, he was building a railway maintenance product that works with Network Rail, who's the railway infrastructure owner. So we basically joined forces to launch a software product for Network Rail that would manage all of the assets across the uk. So for the average user of the railway, like myself, I get frustrated when my train doesn't arrive on time or there's a delay, uh, and I arrive a little bit later. 90 to 95% of the time, that is because an asset, a physical piece of the railway has gone offline. So it could be a signal box, it could be a piece of railway itself, meaning that it's unsafe to travel at a specific speed. Now, that really has an impact on my day and I could imagine has an impact on hundreds of thousands of other travelers. So by maintaining and understanding where those problems are going to occur, you can send teams out when they're needed to fix them. Obviously, if that happens during the middle of the day, it's going to cost hundreds of thousands of pounds in refunded tickets and obviously Network Rail, hundreds of millions of pounds per year. So we built that product for Network Rail, we launched it, it was used across the UK by Network Rail, um, team members. And during that period of time, we realized that other industries had a similar problem. They were using spreadsheets to manage a large volume of Operations, uh, there were errors in their spreadsheets and they couldn't really visualize it or connect them all together. So they lacked that relationship between all the different data points. So they had a lot of information but struggled to make decisions really quickly and they had no ability to use AI tools. So we took the risk of starting our own venture in this space, going bigger than we ever gone before. Instead of winning a contract and building something against some requirements, we decided to build an all singing or dancing platform. So with Mapify, now even I use it for my own personal logistics. So whether it's maintenance for my flats, uh, or you know, traveling abroad, there is a cost element to all my logistics and there is a revenue element. So even as a small individual user I'm using it for a small fee per month. And then we have users that are doing anywhere from 200,000 to a million pounds month in revenue and they're moving large volumes of goods. So some are really critical. So they could be powering biogas facilities and if they don't get the waste or input material then there's no gas that comes out the other end. So it depends on how you want to use Mapify. And what we discovered as we started to build it was Mappify is becoming an operations or an operating system for any logistics across the uk, across the world and we have users everywhere. So the name Mapify, funny enough, came from one of our first user interviews where one of our users just said, what I really want is to be able to see a map of all my assets and know what's going to happen when and how. So that's what it started as. Can I just map my assets? Can I just Google maps my assets?

Speaker A: Yeah, absolutely, absolutely. I'm just keen to understand if there's like a particular profile, uh, uh, a particular uh, target audience prospect that you have in mind because from what I understand and what we've learned so far, you really can tackle so many different industries.

Speaker B: So simply put, we work mostly with warehousing and logistics operators. So they are generally, uh, our icp, if you were to call it our idea, customer profile is a small to medium rise. They're generally generating anywhere from a million pounds in turnover to 10, 20 million pounds a year. They have a small team, uh, maybe a small team of let's say sub five people running the operations of the business. If there is any downtime or any issue relating to an asset, be it a vehicle or forklift, it has a, it has a real impact on the business that day so they might lose anywhere from, you know, £10,000 for a missed delivery, all the way up to £100,000. So there's a real impact on their operations. And they're traditionally using spreadsheets or pen and paper or it's sort of off the cuff, sometimes sending WhatsApp messages and images to kind of communicate. So we work with them closely because they understand the problem, there's an immediate pain on their business. But as you said, we've started to onboard users from across the uk, across different industries. We've had property managers, we've had personal users like myself. Uh, we've had a whole range of, let's just call them industrial, uh, facilities who want to use our products to manage their assets, be it manufacturing equipment, packaging lines, that sort of thing. So we initially started with that ICP of warehouse and logistics and soon we'll be launching a version of Mapify that anyone could sign up to and you could get started for free. So we're willing to prove the value to you by allowing you to use the product free of charge, uh, and start mapping your entire enterprise. And within the first day or the first couple minutes of using it, a lot of our users immediately want to do more.

Speaker A: Keen to understand how you actually, like, use this type of technology. You know, I'm going to refer to the most simplistic way of saying AI ultimately, you know, let's say we've got, uh, a forklift truck as, as you've been referring to. Let's say it's broken down or maybe, you know, a lorry on the motorway, for example, which is, you know, transporting goods. I'm keen to understand how Mapify, uh, actually is aware of this type of data. Is it a case that someone needs to manually input that data there and then there will be data, uh, you know, surrounding that, giving it context, you know, what's the potential revenue loss for something like this? Um, how can someone mitigate maybe that risk there, or what's kind of the protocol as to how we can maybe solve this particular problem? M At this point in time, I'm, um, just keen to understand and how Mapify ultimately works there.

Speaker B: Sure. So I'll give you a couple examples. Uh, I'll start with, uh, one of our core users. So they use it on a daily basis to manage all the activities that perform across their operation. Of course, when you first start, you upload all your existing spreadsheets, you know, a list of assets, a list of personnel, a list of regular activities you perform, and then in the Platform, you can just recur those activities. So you may have deliveries that occur every Tuesday and there's a certain revenue for that uh, delivery. It occurs on a certain time. Generally it's performed by the right supplier. So you may have suppliers in your organization or team members performing that activity. So that's the first step. And then every day as you're using Mapify, either you can use it on your phone, a tablet or desktop, you'll get notifications or you'll get a pop up that will show you if there's any conflicts that day because really you need to know if there's an issue, otherwise things are smooth uh, and everyone's happy. So it will notify you if there's a conflict or something has not happened and then you can resolve that conflict. It might suggest to you that you call the supplier to reorder or to rearrange delivery or in a larger case, let's say as you said, a truck is broken down because you can share Mapify with all your suppliers and your customers even though it is your organization, all viewers are free means um, that if there's breakdown instead of it going from the driver to the driver's manager, from the driver's manager via email to your logistic manager to then go to the technician or the person receiving the goods on site, which is four different emails, it might either never arrive, will arrive too late. So with Mapify everyone sings off the same hymn sheet. So that driver can just go into the platform, mark it as you know there's been an event, the uh, delivery won't arrive and then immediately they're able to dispatch something new. Not just on the supply side but also logistics site. So that's one example to uh, give you one from my own context. I, you know I'm a property owner, I have a flat and my flat has maintenance that's always due to. But as you may imagine the uh, bank account is not limitless. There is things that you can only certain do in a uh, given timeframe. So let's say you want to replace your shower or you need to replace your boiler or you know, God forbid there is a roofing issue. Now uh, even as an individual user, when I started to see my customers using Mapify, I immediately realized I could use Mapify for my own use. So I now can plan ahead and say oh, we need to replace this part or I live in a stairwell where there's 10 other landlords, 10 other flats, we need to coordinate together to replace parts of the roof or there is a renovation that's needed to the stair. We need to replace lights. So I started using it because there's too many activities, too many assets, too many moving parts. And actually I need to plan the investment with the costs that incurs for me to have a property. So that's a really small. And I've shown you a really big case. And uh, on the enterprise level you've got people like Network Rail with millions of assets spread across a geographic area, almost impossible to handle as a human being.

Speaker A: Yeah, no, I, that, that makes, that makes complete sense. And I like products that do make sense to me, um, because I'm not the most technically apt. Uh, I guess that brings me on to the platform itself because a few discussions on this podcast have encompassed, um, user experience, user interface, um, how intuitive a product is. Um, because of course we have to cater for the non technical stakeholders, we have to cater for the non technical users. I would imagine that user, uh, interface, user experience, ah, that whole realm of things is going to be incredibly important for Mappify and its application given the type of user, uh, um, you know, potentially working with the app.

Speaker B: Yes, that's true. I think the, the challenge with a lot of products is how easy it is to use, not just to adopt and to onboard, but also to use going forward. So we are quite focused with our users on improving that. So we release new features every week. So if a user says to us, oh, it would be really nice if I can filter by X or I'm, um, finding it really frustrating to constantly copy some of my activities. So within that week we'll release a new feature that allows them to quickly filter by activities that are overdue or to create a new dashboard that allows them to see events that are open or they've been open for more than, let's say 10 days and haven't had a resolution. So because we can move quite quickly, again thanks to AI tools, we can write code much faster. There are open source components we can use and what we've started to do is work with the users to cut parts of the process that add perhaps an extra layer of complexity. So you don't need to see all the fields when you're submitting an activity or um, when you're sharing with your team members, you can, you have some control, but you may just want to quickly share a viewer license. So a lot of the time instead of trying to build something perfect and design it, you know, in Figma or on a piece of paper, we work hand in hand with our users and they tell us what they want and we iterate really quickly with something really simple and then improve and improve.

Speaker A: Yeah, yeah, okay. I, you know, I'm going to relate uh, this back to an app that I use daily as well is Runner. And I really like the fact that on Runner it has like this type of knowledge base. Um, and actually uh, you can speak to um, you know, the team over at Runner, uh, you know, actually suggesting, um, you know, new uh, features, um, ways in which they can optimize the application. Um, and I like it when businesses are present in forums, whether that be the knowledge base itself or potentially Reddit. So I think you're doing a really, really excellent job in actually working, uh, very closely with your users, making them feel valued, making them feel heard. I think that's a, ah, really, really nice touch there.

Speaker B: Um,

Speaker A: moving forward, what does the future of logistics look like to you?

Speaker B: To me, logistics is actually quite a simple industry, but it's very complex because there's lots of moving parts, there's a lot of people and a lot of decisions need to be made. And a lot of the systems in logistics are disconnected. So for example, uh, we recently onboarded a user who is moving biological samples around the world to conduct some tests for a scientific purpose. So let's say a clinical trial. And that should be fairly simple. You know, you order something from a part of the world, you have a tracking link for that delivery and it arrives. But there is a lot of documentation required, everything from customs all the way to ethical documentation. You may have to do some sort of assessment from a safety perspective on what you're moving. And that's where the complexity comes in. A lot of this can be automated. So for example, with Mapify, you can attach all that documentation to that activity, that movement, including the tracking length. And in the future what we imagine is that all of the operating systems would interact with one another. Now thankfully, we have AI agents now and these AI agents can learn from one another and interact with these components. So I don't have to build a custom API every single time. We have an AI agent that can go out and interact with that platform. So one day we imagine perhaps Mapify at the center of it, uh, an orchestra of different agents and systems interact with one another to perform that task. So if you were a, uh, small enterprise, let's say in the scientific space, and you wanted a sample, you would just say to your agent, this is what I would like. Or you would just say to Mapify, I need to order this. And it maybe Knows from your previous history you've ordered one of those before, it would interact with that supplier. Send them an email, make sure the quote matches your approved budget. Orchestrate the entire delivery operation and you just have something arrive at your door. Sort of the way that Amazon does it for us as a retail customer is, we imagine the same thing, logistics. And it can also apply to the warehouse. So you've got, you know, a number of assets that are critical to your operation. A JCB or you know, some sort of forklift that without it you cannot perform your daily activities. It should already be maintained before a driver even gets that vehicle. And the suppliers involved in maintaining or servicing that vehicle will be constantly in the loop through mapify or a platform like it. If something changes or that forklift needs to be moved to another warehouse, automatically that maintenance schedule will be rescheduled. So a lot of the things that human beings are not good at, sometimes juggling, you know, spinning plates on top of spinning plates, that can be done really, really well by a mission. And that's where AI I think, has uh, superpower above our abilities at the moment.

Speaker A: Sure, sure. And just to uh, give a slight nod to the, to the map, there's a, you know, um, uh, I guess a very influential and fundamental part of, of the platform itself. Tell us a bit about, you know, where you're getting this, uh, where you're getting the map based data from or how you're actually putting this map together. Yeah, you know, is it like geospatial data that, that you're, that you're, you know, resorting to there? I'm m keen to, to learn a little bit more about that.

Speaker B: Very Simply, we use MapBox. So we use an existing tool. No point in reinventing the wheel. And thatbox provides the layer that the user sees. We overlay the locations of the assets. Some assets for example, might be regularly moved. So let's say you've got a delivery operation. It leaves a warehouse at a certain time and arrives at a customer site at a predicted time. So you've got the movements already based on that delivery. We can integrate live tracking and live sensor data. But from a mapping perspective, most people just like to see their assets or their activities on a map. Oh, I need to perform this many deliveries today. Or as an example, I have maintenance activities across this network and I only have one allocated personnel, one technician today. Is it possible? No. And that's where the map shows that to you visually. Obviously we have a conflict alert that comes up and says you've got One engineer today, they're not going to perform all these activities. We suggest that these three activities are prioritized because they generate you know, 90% of your revenue. These other three are non urgent. They have no compliance requirement against them, therefore they can be postponed till tomorrow. So these are the types of things that a machine is very good at. And a map you can visually see, this is red, this is amber for example, this is green. So with one view, one click mobile or desktop you can see what your operation is looking like. A lot of people just want to put simply to them. They don't want anything complex, they don't want a huge description, they just want bury and say what do I need to do today? What asset has spare capacity? These are the questions they ask. Well, how can I grow my business if I'm constantly stuck in the weeds dealing with Excel or trying to administer my operation? I don't have the time to grow it.

Speaker A: I think it's, I think it's excellent. I like the use of uh, well the introduction of like an LLM, um, like an AI agent, you know over the next two weeks there. I like the fact that it ah, makes suggestions um, based on, on priority and compliance and profitability. Um, I like the fact that you've got the uh, map based data there um, to give the data context. Um, yeah, I think it really is a very, very interesting business and you guys are in your infancy right? I'm really, really interested to see you know where you are going to be in 12 months time. Tell me, you know, you don't necessarily need to spill too much tea but um, where do you see yourself in the next 12, 12, 18 months there.

Speaker B: Uh, as you said, I think we are very young. I think the company realistically was only about 6 months old. We founded in April of this year. Obviously we've had a lot of experience in the sector so we've brought that to bear. In eight plus years working in logistics and manufacturing. What does it look like in 18 months ahead, if all goes well, we will be able to orchestrate your entire business operations within the logistics context of course from the Mapify platform. So that doesn't just include asking questions via search or summarize or reporting on what's already in your operations, but conducting emails on your behalf based on a set of rules. So if a delivery is, is late or it's over June hasn't been marked complete, you know, sending your suppliers reminders, that's a really simple one. Uh, on the other end in 18 months time we could also integrate all of your existing business solutions into Mapify. So for example, every time you send a receipt or an invoice that could be automatically attached in Mapify so you can see in one place, that's the simple things. And then if we really push the boat out, if we were to say what would that look like in the future? We would be able to automatically predict when something significant is going to happen to your operations or all the risks that could occur. For example, you want to do something really big like take on another million pound contract, you know, there is an impact that will have on your organization. You may have to hire new people, it may have to change the way you operate. And that's where Mapify can step in and say, you know, by winning this 1 million pound contract, you'll need to purchase a new asset. We think you should purchase this type of asset based on the downtime of these other assets. And we can bring to bear our knowledge across different industries and different customer profiles to say we think this is the best decision for your organization. So we can not only help you to grow your operation operationally, but you can also invest better, have a higher profitability rate of return and you can expand your portfolio all without having to increase your profits.

Speaker A: Amazing. Amazing. Tell us, is um, Mapify hiring at the M moment? Anyone uh, interested in joining the business? Uh, are you looking to potentially expand the business in particular areas over the coming months?

Speaker B: We will be looking to hire uh, not for this quarter but come Q1, Q2 of next year we'll bring on um, some new engineering team members as well as some account manager scale. So our biggest challenge at the moment is that we've got an extraordinary amount m of demand. It's lovely but it means that we spend a lot of our time working with users. So soon we'll be able to new types of users, bigger companies and we'll need additional development talent for that, additional sales or commercial talent. We can work our customers through that journey, understand their needs and communicate that back to the team as we need to develop.

Speaker A: Amazing, amazing. That's great. That's great. Really, really exciting stuff. I guess anyone who is watching, um, listening, um, do bear that in mind over the next six months. Look out for um, you know, expansion there with Mapify. Um, so no, that's, that's really, really exciting. Well look, um, I, I think that that brings us to a close there. I, I, I've really enjoyed the conversation today, learning more about Mapifying. It's definitely one to watch for me. You know, a great deal of potential and it, you know, does answer real world problems. Um, I love the fact that, you know, enterprise users, uh, can use it there and then individuals like you and I, um, can also, uh, you know, advantage from that. Um, so, uh, yeah, no, really, really appreciate you joining me today on today's episode of, uh, Earth, ah, observed there. Uh, um, yeah, we'll uh, well, we uh, we look forward to hopefully inviting you back on over the next six to 12 months to see what the updates are and, and see where you are then.

Speaker B: Thank you for having me.

Speaker A: No problem at all. Thank you very much.

Speaker B: Cheers.

Speaker A: Hello and thank you for watching today's episode of Earth, uh, Observed, powered by Deepreck AI. If you're interested in joining me for an episode of your own, uh, to discuss a topic of your choice, do get in contact with me. And of course, if you are looking to scale your team within geospatial Earth, uh, data or atmospheric AI, please do get in contact. I'd love to understand how I can support you. Until next time, have a great day.