Taylor Shropshire is the founder and host of FishCast, a podcast dedicated to marine conservation and fishing. On this episode he tells Tom how he built FishCast from the ground up, the technical mistakes he made with his early audio setup, and why he believes storytelling — not catch reports — is the key to getting anglers to care about fisheries management.
Press play above to watch, or scroll down to listen to the full conversation.
FishCast is a podcast founded by Taylor Shropshire that focuses on marine conservation and fishing. The show features conversations with anglers, scientists, and conservationists to tell stories about fisheries management and connecting people to the resource.
Taylor Shropshire is the founder and host of FishCast podcast, a marine conservation and fishing enthusiast who uses podcasting as a platform to connect anglers with conservation stories and fisheries management discussions.
Starting a fishing podcast requires good audio equipment, recording software, and a clear mission for your show. Taylor discusses the specific technical setup and gear choices that matter most, and why defining what makes your show unique matters more than chasing production value.
Podcasting allows for long-form storytelling that connects anglers emotionally to conservation issues and fisheries management. Taylor believes this format creates deeper engagement than short-form content and bridges the gap between anglers and the scientists working to protect fish populations.
FishCast focuses specifically on marine conservation and fisheries management rather than just fishing techniques or catch reports, amplifying voices from scientists, advocates, and conservationists alongside anglers.
I like talking to people who are building something in this space the hard way, and Taylor's done exactly that with FishCast. He didn't start it because the fishing podcast world needed another show about catching fish — he started it because he saw a gap in how conservation stories get told, and I wanted to hear how he built that from nothing.
Taylor didn't set out to make another show about catching fish. He saw a gap in storytelling around marine conservation and wanted a platform where anglers could hear directly from the scientists and advocates working to protect the resource. He talks about the moment he realized most anglers don't understand the complexity of fisheries management, and why he believed a podcast could bridge that gap. Worth hearing him tell it.
Tom and Taylor get into the nuts and bolts of production — microphone selection, recording software, the editing process most listeners never think about. Taylor shares the specific mistakes he made with his early audio setup and what he learned the hard way about sound quality. If you've ever thought about starting a podcast, this section is worth a close listen.
Every host has episodes that stick with them, and Taylor opens up about the specific conversations that changed how he thinks about fisheries management. These weren't just informative interviews — they were perspective-shifting conversations that made him a better angler and a more informed advocate. He names names later in the episode.
What struck me most about this conversation was how thoughtful Taylor is about the role of media in conservation. He's not just creating content, he's building a bridge between anglers and the scientists and advocates working to protect our fisheries.
We got into the weeds on podcast production, and if you've ever thought about starting your own show, there's real value in Taylor's experience — he's made the mistakes and he's generous about sharing what works. If you care about the future of our fisheries and the role storytelling plays in conservation, listen to the whole thing.
Taylor Shropshire · FishCast · Fathom Science · Tom Rowland
Taylor Shropshire is the founder and host of FishCast, a podcast dedicated to marine conservation and fishing. Through FishCast, he creates a platform for conversations with anglers, scientists, and conservationists, telling stories that connect people to the resource and explore the complexities of fisheries management. Taylor is head of ocean analytics at Fathom Science, an ocean and weather forecasting company, and is a lead developer of the FishCast offshore fishing forecast.
Tom Rowland Podcast — Episode 998: Taylor Shropshire
In this episode: how FishCast puts an AI fishing forecast on your Simrad unit, decades of daily fishing reports that trained the model, why he calls himself the Jonah Hill of Moneyball for a fishing fleet, a $100-for-the-week subscription you can try before The Bahamas, and a conversation that goes all the way from fishing to quantum computing — in the exact words spoken.
Tom Rowland: If you want everything you need to take care of your boat, StarBright's got the solution. It's called boat care in a bucket. That's right. One three and a half gallon bucket packed with essentials, vinyl bright, deck cleaner, marine polish, a wash mitt, hull cleaner, and boat wash. That one kit will keep any boat clean and protected, but that's just the start. I also use StarBright's salt off, which is incredible for rinsing salt off after a day on the water. Their brushes and cleaning products work just as well on your RV and patio furniture and anything else that lives outside. It is a company that gives back. That's one of the things that I really like about StarBright. StarBright supports marine conservation through Project Sea Safe, and that means a lot to me personally. You're getting performance and purpose. Visit starbrite.com. That's starbrite.com, and get your boat dialed in for the season. From The Seychelles to The Keys, I trust Nikon binoculars to find fish fast. Visit nikonusa.com and upgrade your line of sight. I thought I was covered, but I wasn't. HH Insurance showed me that my old policy did not cover The Bahamas or nighttime navigation, two major gaps that could have cost me a lot. Jake and Landon fixed it fast. They know charter policies better than anyone I've ever worked with. Call (727) 498-5551 or visit hhinsgroup.com and tell them Tom Rowland sent you. Trusted enough to go with me to The Seychelles, my Danco pliers are a staple in my kit. Check out dancopliers.com and build yours today.
Taylor Shropshire: My name is Taylor Shropshire. I'm the lead developer of Fish Cast, and this is the Tom Rowland podcast.
Tom Rowland: Taylor, what's up, man?
Taylor Shropshire: It's going well. How how is it on your side?
Tom Rowland: Good. Good. So you just got back from, Fort Lauderdale Boat Show. Is that right?
Taylor Shropshire: I did.
Tom Rowland: Yes, absolutely. I should have, I I thought we were going to, maybe run into one another there, but we didn't. So it's nice to finally meet you and put a face to a name. About what you just, unveiled at, Fort Lauderdale Boat Show with SIMRAD.
Taylor Shropshire: Sure. Yeah. I can give you a little background to you. First, so I'm head of ocean analytics at a company called Fathom Science. As I mentioned, I'm also a lead developer of Fishcast, which simply put is the world's first multiday offshore fishing forecast. So as you mentioned last week, we were at the Port Lauderdale Boat Show. We're releasing the newest version of, Fishcast called Fishcast Powered by ROS. And for those of you that are not familiar, ROS is a company that specializes in offshore fishing reports, and the product is exclusive to Simrad. So that's why we were there last week.
Tom Rowland: And so how or is it available on any Simrad product or on certain ones? Or
Taylor Shropshire: Yeah. It's, it has to have the latest software, which they they call their Neon software. So NSS four is the one that they're promoting, But it essentially just has to have that Neon software, which is the newer one that will have for all of the NFTs moving forward.
Tom Rowland: Okay. And and this product is designed mostly for offshore fishing. Is there are there any other applications to it?
Taylor Shropshire: No. It's purely, offshore, fishing. It's so it was developed based on the idea that anglers should have access to a forecast just like they do for weather.
Tom Rowland: Okay.
Taylor Shropshire: So, you know, the idea is you should be able to track water, see what the conditions are looking like, this weekend, kind of plan out your trip. There's, you know, lots of fishing apps out there that use, like, satellite data. And so we really saw a need for creating something that people could use as a planning tool, and ultimately just help people enjoy offshore fishing more.
Tom Rowland: And so as a planning tool, can you is there an app that goes with it or something like that, or do you need to get on the boat?
Taylor Shropshire: So it's a subscription data service, potentially exclusively on some of our devices. It doesn't require any extra hardware. You just download the the today's forecast before you leave the dock. And, the forecast contains, a three day forecast with three hourly intervals, so you can watch how the water moves over the next three days while you're out there. And it contains three data layers. So the first and most exciting is what we call our optimal catch location map. And essentially what it is is a data analytics fishing analytics, data layer that shows you high and low probability areas for finding large game fish, like billfish, tuna, wahoo, banyi. And then I like to say we know that analytics will never fully, like, replace experience on the water, so we also provide two additional data layers. One is ocean temperature data layer, and the other one is ocean color. So those are both three day forecast as well. Helps you to identify, water mass boundaries and track source water.
Tom Rowland: So somebody on your team had to be, avid fisherman. Is that you?
Taylor Shropshire: Yeah. So, I mean, I would not describe myself as a seasoned offshore fisherman, but I grew up doing a lot of fly fishing. Definitely done some saltwater fishing. So I'd give you a little background about the company, which might make a little bit more sense.
Tom Rowland: Sure.
Taylor Shropshire: So, Bantam Science is an AI tech company based out of Raleigh, North Carolina. We specialize in developing ocean forecasting and weather forecasting systems for ports around the world, global shipping companies, offshore energy platforms, and most recently, offshore fishing applications. So the company was founded by a group of researchers in the oceanography department at NC State University, myself being one of them. So I have a PhD in oceanography, which I like to say means that I might not be great at catching fish, but I do not define them. And that's exactly, what Fishcast is designed to do.
Tom Rowland: That's cool. So the a lot of the data that is in this, it sounds like some of it is available, like, like the wave heights and the forecast and things like that that you're probably pulling that from somewhere. But then others is more like that. It sounds like that's what a fisherman knows. It's like how how the wave height will affect the fish or the temperature is gonna affect the fish. Yeah. So how do
Taylor Shropshire: you how
Tom Rowland: do you go down that road of putting the human touch on it?
Taylor Shropshire: Yeah. So, I mean, like I said, we specialize in ocean forecasting. So we're we're developing these models, computer models. The best way to describe it is, you know, like, when you turn on the TV and you see what the wind forecast is gonna be for today, how fast the wind's moving, they're using a computer model to predict that. We're doing the exact same thing but for currents and for other oceanographic variables. So we do all the ocean modeling in house. That's what our team specializes in. But the human element, the part that you just mentioned, comes from, our partnership with ROS. Are you familiar with Ross?
Tom Rowland: Not really. I'm not that much of an offshore fisherman, but there's a lot of people that are listening to this podcast will be. Yeah. But are they the one that you subscribe to and and that you're getting the forecast up to the second in on your vessel?
Taylor Shropshire: They've they've been around for decades, and they use satellite imagery and write detailed fishing reports every day for, like, 60 or so regions along The US East Coast, Gulf Coast, and Bahamas. And they've been a customer of ours for a few years. You can imagine one of the challenges of satellite imagery is sometimes you have clouds, and nobody, you know, wants to pay for a fishing report that has an image, a satellite image with a bunch of clouds in it. So we provide them every day with a ocean forecast, from a model, so it's cloud free. And a few years ago, into our partnership, I asked them, hey. What do you guys do with those historical reports? And they're, like, like, nothing. We, you know, just have them on a hard drive basically somewhere. So I asked them if they could share those with us. So they they gave us, like, 10,000 detailed phishing reports for The US East Coast, Gulf Coast, and Bahamas, And we wrote, some software to basically read through each one of those reports, pull out all the fishing recommendations that they that they, put in there, and develop a model that basically identifies the most optimal conditions. So would
Tom Rowland: would you also when you take all that those 10,000 reports or how many ever you said, you you take all those reports and then do you go back through historical meteorological data and see exactly what the wind was doing?
Taylor Shropshire: And, man, that's incredible. Yeah. You're hitting it right on the mark. So our ocean models, they go back to about early 1990s because that's when satellite data really started to become more available. So what it means is that in the ROS reports, they, specify specific, like, waypoints, like the lat long of, suggested fishing locations that they have identified. They're expert oceanographers. They look at satellite imagery every day. They're getting calls for people on the water. So we use those points, go back in time, and say, what were the ocean conditions like at each one of those fishing hotspots? What was the temperature like? What was the bottom depth like? What was the currents, the salinity, sea surface height, tides? We pull out all that information and basically build a machine learning statistics kind of AI model to identify, the most, optimal conditions for for fish.
Tom Rowland: That's really incredible. It's it's funny because, like, when I was thanks, Moose. Moose is sniffing the microphone. When when I was a fishing guide, what we would do like, what someone told me, you know, on the inshore side of fishing, my my mentor, Simon Becker, he told me, like, I was I was trying to take down all these notes about every single thing on the day. And and it's just too much writing. You get tired. You get home at the end of the day. You're tired. You can't you can't keep that much data or I couldn't write down that much data. Like, I wouldn't mind jotting down a few notes here and there. He He was like, look, all you really need to know is what time you caught a fish or what was going on that you wanna replicate at a certain time and what the date was. And that's really it. And I was like, how is that it? He's like, well, there's published, you know, NOAA Yep. You know, forecasts about what it was. Or you could go out and you could look at the the light data, you know, at a at a at a like, we have Sankey Light or any of these lighthouses, they'll have data there. And so if you knew that you caught a fish at 12:00 on this particular day, you could go back and you could see exactly what the tide was doing. You could see exactly what the wind and and, seas were doing and what the temperature was and all of the factors that are there. But it was so manual and so slow. And mostly, you could maybe get one or two spots really, really nailed down like that. And if you really, really, really worked at it, you could maybe, you know, start to understand a little bit of a trend. And but more than anything, you were just doing it on experience. And this was kind of a a loose guide to going back out there and seeing if you could replicate the same thing. But what you're talking about is really so different with machine learning and AI. And I'm fascinated with AI, and I'm scared to death of AI. Yeah. And I use AI every day, but then I also think, where are we where are we going with this? Yeah. And this has be a really interesting conversation with you who has developed all this, and this is this is your business of AI. Where do you think that we're going with with AI as it applies to fishing first, but then just overall?
Taylor Shropshire: Yeah. I mean, I'm the first to say, okay, I was not a silver bullet, and I think, you know, it definitely has some negative connotations with it and maybe overhyped in some aspects. You know, for for phishing, the way I think about it, at least in offshore fishing, there's really three steps to the fishing process. First is finding the fish. Second is having the right rigs in the water, and the third is fighting, you know, the fish. And no one wants to spend the whole day doing the first. Right? And so, I think there is a fine line of how do we give people information that helps them enjoy being on the water more, but doesn't remove that spirit, so to, you know, so to say, of being outdoors or or, you know, being more observant and kind of learning the water. So I think AI definitely has a a role just in helping us enjoy the outdoors more, but there I think we do have to be aware of ways that it could, you know, not be used so well. So I'm I'm definitely not one that just blindly uses artificial intelligence and say, oh, this is gonna change everything, and we're just gonna not ask any questions. So, yeah, I guess the answer is I think it's gonna be a tool, but not, you know, not a silver bullet.
Tom Rowland: Yeah. When it comes to phishing and and replicating, I mean, that's really what we're trying to do, is everybody kind of comes across something. And I'm sure even if you're using these hotspots or or your fish cast, maybe you deviate from that. You're like, well, this is supposed to be working here. Let's go try something else. And you're leaving from a place that you wouldn't normally be, and you're running to a place that you wouldn't normally go. And in between that is some water that you don't normally see. And you come across something that is amazing. Big bait ball, something floating. I don't know, birds diving, and you have a great day. And then you start thinking, okay, why was that happening? Was it luck? Did we just get did we just come across something that we don't normally come across, or was there something happening there figure out, okay, on this wind, on this tide, on this these conditions, we could maybe have a chance of replicating that in this water depth and this, you know, all these factors coming together. And that seems to be what AI seems to be Yeah. Really, really, really good at is taking data, understanding that there's there are trends here that maybe maybe we're not able to see unless you really, really study it or team for studying it.
Taylor Shropshire: Yeah. That's exactly it. I mean, the fish are gonna be in a specific habitat. Right? They have preferences, temperature, water depth, right, exactly as you said. And for us, we can kind of scan the water user experience. But when you're using data layers, multiple data layers now, and you're asking AI models, say, show me the most optimal conditions, where where are they for today, that's a really powerful thing to be able to do. And you like like I mentioned in our service, we we provide a data layer, called optimal catch location map. So it, is essentially like a colored map where red indicates high likelihood or high probability of finding large game fish. So you can look at it and with the temperature and ocean color forecast and see, does this make sense to you, you know, based off your experience? Or maybe you try you know, go try something and it doesn't work. You can use this as a tool. Okay. The analytics say that this is gonna be a high, you know, high probability area. You can't ever say for certain anything, but you can't argue with statistics. Right? It's this is historically the conditions that have been favorable for fishing, and that's what Fishcast provides.
Tom Rowland: Wow. So on all these different, kind of technological advances that we've had that get get, kind of assimilated into the electronics, like what what your product has done. I mean, it happened first with kind of, you know, a base layer map, and then it happened with, you know, to where we're now, you know, Google Earth. You can get on your on your GPS, or you can you can have a really high definition photograph or or and then you can see your boat actually running across this photograph, and you can see things like individual mangroves and, like, crazy things on there. And so when when some of this stuff happens, you'll have different, kind of, adoption from different levels of anglers. For example, some of it, you you know, a very, very experienced angler looks at something and they go, well, that's that's that's good. That's just kind of for forget beginners. Right? Like, it it happened when when the there would be, like, on the sonar, it showed people how to read read the sonar a little bit, and it showed, like, a fish icon instead of an arc. And so a really high level fisherman is, like, no. I wanna see that I wanna see the arc. I know how to read that. Right?
Taylor Shropshire: Right.
Tom Rowland: But somebody that just gets started is looking at that and going, oh, I never could tell the difference between the bottom and an arc anyway. So that was kind of a beginner function and really helped some people get into the sport and help some people to find fish. But more advanced anglers weren't interested in it at all. So I'm just kind of wondering with your product here, do you see that as a product tailored for the super experienced captain that's gonna help them, or is it for the beginner, or is it really just an advance meant to the entire sport?
Taylor Shropshire: Yeah. I mean, I think it's the latter. I think it's for everyone. And the reason why I'd say that is because, as I mentioned, we have that optimal catch location, kind of fishing analytics map that shows you high and low probability areas. But then we also have the ocean color and temperature forecast. So if you really know what you're looking for, you're looking you know what temperature water you want to fish for the day, you want to target a certain break, you know, the blue to green break, you're gonna be able to use that forecast to see how the water is moving. And really, that has not been possible before. There's been other services out there, that use, as I mentioned, lots of satellite data and provide some information, but really to have a forecast for a planning tool, you know, to be able to be at work and look at what what are the conditions look like for, you know, this weekend, that I think that can appeal to beginners and expert anglers.
Tom Rowland: So would it do something like tell you the strength of the of the current, like, offshore? Because a lot of times that, you know, that's what you run into is like you're you you know where to go. And the winds blowing one way and the currents going another way, and you end up fishing right up your anchor line. And you probably if you had known that leaving the dock, that's probably not the place that you would have gone. Is is it kind of giving you that data?
Taylor Shropshire: So far, it really is just three there's those three data layers. So it's the analytics, the the, temperature maps, and the ocean color. So that's all it's all dynamic, so you can see it moving. You can see the forecast every three hours where how the how the front is changing. You can see that for the next three days. There's a play button at the bottom of the MFP. Just click play and you can just watch how the water is moving. And there's 133 different inletsmarinas that you can pick from. So it's really high detail ocean forecast for your specific inlet. And analytics are all specific for your specific inlet as well. So currently, it doesn't have ocean currents as far as, like, a separate data layer. But ocean currents go into the analytics, trying to keep it as simple as possible for people to use. They don't have to, you know, look through tons and tons of different data layers. Three data layers, analytics, temperature, ocean color, gives you all the information you need about the conditions and where the high probability areas are.
Tom Rowland: What's been the feedback so far?
Taylor Shropshire: Yeah. Fort Lauderdale is great. Lots of excitement. I've had a couple of people say it's really, intuitive. You know, it's easy to use. It doesn't have all those, data layers you have to turn on and off to try to figure out what you're looking at. I think, you know, it's gonna be like anything that has to build confidence in the community. And that's one of the questions we get pretty quickly is what's the accuracy, you know, of this? And, kind of answer it in two different ways. One is we've had some beta testers, handful of professional tournament offshore fishermen use it. We've had really good experience with it. The other way that I'd answer that question is we're using Raf's, historical phishing recommendations. And they've been doing this for multiple decades. So you're basically getting the cumulative experience of decades of people who are who have all they do is look at satellite imagery and pick phishing locations in fish casts. So we have high high expectations for it to become a, you know, widely used product.
Tom Rowland: So what about, what are the what are the regions where this is working? Like, just where they've been putting these phishing reports out for the last
Taylor Shropshire: Yeah. So so this particular version of Fishtasks, Fishtasks powered by ROS is just covering areas where, ROS has phishing reports or approximately that area, which is the entire US East Coast, the entire Gulf Coast, and The Bahamas. So as I mentioned, there's 133 inlets that you can pick from. So you get about a 120 mile radius around that inlet of data every day that you download right right when you get on your boat. And then you have a three day forecast that your hand works like at your fingertips. Compared to other products, that are on the market right now that only update, you know, twice a week sometimes or once a day or giving you three hourly, information while you're on the water.
Tom Rowland: Layered with weather, like a standard weather, like XM? I know the XM rate weather is available on
Taylor Shropshire: Yeah. So this is fully integrated in the SIMRAD, MFP. So any data layers you want to look at, the high resolution bifimetry, you can overlay that on top of the fish cast data layers. Yeah, you can look at weather on top of it. They're they're full colored images. So you can imagine you couldn't have, like, both two colored images on top of each other because you wouldn't be able to see it. But, there are some layers that have kind of, like, a transparency with them so you can overlay multiple layers together.
Tom Rowland: And so out of all these people that have been testing it, have you had some success stories?
Taylor Shropshire: Yeah. Yeah. We have had some success stories. As I mentioned, we've had a handful of professional anglers try it out. Rafsan, in particular, has a lot of clients that that use their services that, you know, do all the biggest tournaments. And, so they've gotten some feedback directly from them. Navico has some of their own kind of, you know, brand ambassadors try it out. They've been really pleased with it. Yeah. Ultimately, I think everyone has to to experience sort of themselves to to see, you know, the quality of the data. But we'll we'll put our data up against anybody. We came from, you know, academia, NC State University. All we did were was develop ocean models for years and years for scientific, study. And now we're using that same really high quality models to to basically provide, oceanographic data to a range of different industries.
Tom Rowland: Wow. And so, with the with the fishing, application that you've done, what what how do you kind of use the same approach for the shipping industry, or do you do do AI kind of powered things for other
Taylor Shropshire: industries? I mean, the ocean is one of the most data poor areas, in the world. And you can imagine it's just it's so much more difficult to get observations, in the ocean compared to on land. So, really, our bread and butter, as I mentioned, is just is developing ocean forecast models. So predict what the currents are doing, waves, tides, other oceanographic variables. And that information is just so valuable for making better decisions on the water, whether you're doing fishing, whether you're, you know, taking a huge container vessel across the Atlantic, you, you know, you you can understand why currents would be really valuable. So it's it's really the same data and the same technology. It's ocean forecasting, AI powered ocean forecasting that we do. It's just using that for different applications.
Tom Rowland: Mhmm. Have you looked into other, kind of data gold mines like you found, or, you know, worldwide? Like, does Australia have, like, this That's a the
Taylor Shropshire: same type of training? That's a good question, because we have aspirations for making FishCast global. Right now, we if you go on our website, fishcast.fathomscience.com, there's three versions that we're currently working on. So the first is FishCast powered by ROS, which is what we're talking about here today. We also have FishCast commercial. So you think the exact same line of thinking. Right? If you're a commercial fishing fleet and you have historical observations of very caught fish, we can create custom analytics, to support that. And then Fishcast Classic right now is exactly what you're talking about is more of, like, a global analytics, for fishing. So instead of using ROS, hotspots, we're using other information that we can gather. So fishy survey information or, just kind of, like, core oceanographic principles that lead to good fishing, like strong fronts.
Tom Rowland: One thing that you just said that, you know, for a fishing fleet. So let's just say somebody has, you know, 30 or 40 boats, and they have been keeping the best records that they can for the last thirty years. You could create basically a a large language model. Is that what you're doing basically with where it's only gonna draw from their observations and not from every other observation? I mean, that's what
Taylor Shropshire: So, I mean, it wouldn't be a large language model. It would just be a they would provide us with, the geographic coordinates of all of their catch locations. And then we would do the exact same thing where we go back on our model, pull all the environmental data out, generate analytics for them. The best way I can describe it have you ever seen the movie Moneyball with Brad Pitt and Jonah
Tom Rowland: Hill? Yeah.
Taylor Shropshire: So for for for your listeners who haven't, seen it, it's a great movie. It would be like baseball. But more importantly, it it shows how statistics and analytics change the game of baseball forever. So instead of picking players, you know, just from experience, they start to use statistics. So I don't like to classify myself as as a nerd in most cases, but I'm Jonah Hill in this case. And instead of picking players, we're picking phishing hotspots. They're using statistics and analytics to show us, okay. These are the conditions that are most favorable for offshore fishing.
Tom Rowland: Yeah. I think it's interesting, though, that that you would do that just for a certain fleet. I mean, I guess for a certain price, you would.
Taylor Shropshire: Yeah. I mean, it it it's really the same technology. It's just what we need are observations of where you had success in the past. And then we can build a model for you that says, hey. This these are the the combinations of conditions that come together in the ocean that have tended to lead to, you know, higher fishing success.
Tom Rowland: Yeah. And how did you get, integrated into into Simrad specifically, but Navico, I'm sure, as you as you continue to go, maybe there'll be other products that should be integrated into?
Taylor Shropshire: Yeah. Well, I think everybody, you know, wants the most convenient way to digest information. And on your MFE while you're out on the boat is just that's the best place to to really be able to work with that type of information. There's tons of apps out there. Right? Everybody's got all these apps on their phones. We really didn't want to be another app. We wanted to provide this data in the most digestible form, which is right on the MFD. And, we spoke to all of the major MFD companies. This is a product that's almost three years in the making. And, Simrad was just the the, the one that moved the quickest and the one that we felt had the best vision for, creating this type of content.
Tom Rowland: Right on. So is it exclusive to Simrad?
Taylor Shropshire: It is exclusive of Simrad, currently. And, you know, can't say too many details, but, you know, I think we like I said, we have great bigger aspirations, to be really the data provider for the recreational boating and fishing industry, over the years. So, yeah, FishCast for us, is really, I would say, our our first step into the space. And oceanographic data, like I said, can be used for so many different applications. Mhmm. So we're really excited to see where we can continue to build with, Simrad.
Tom Rowland: That's that's super cool. I've been using Navico products for thirty years, with started with Lawrence and then, of course, we use Simrad as well. But great great products. I find them easy to use. And and, honestly, people are like, what do you think about this one over here? I I don't know. I don't use them. I mean, that's what we have on our boat. I don't really have a lot of experience with other electronics, but Simrad and Lorentz seem to be very easy to use to me. But I'm a long time user. So I Yeah.
Taylor Shropshire: I've been really impressed with, just their marine cartography marine cartography. They the the resolution that they provide, MFD is just really responsive, because because the data that we're providing, as I mentioned, is very high resolution, and it's able to to to render it, and allow you to to digest it very easily.
Tom Rowland: Do you have competitors?
Taylor Shropshire: So, I think most people are familiar with fish mapping by SiriusXM. And I think, that's a great service. I would I would call them a pioneer in this space. Many years ago, they kinda came out with this on MFP information about, good fishing locations. But I think we're taking a significant step forward with FishCast, to go through a lot of different features. But, the the forecast feature, right, you can actually see into the future of what the fishing conditions are gonna likely look like for the next three days. I mean, that's that's a huge, huge step forward. It's full colored images. So, with SiriusXM, you can use it to look at, you know, like, plankton fronts and temperature fronts, but they're contour lines. So you're not getting the full heat map, to be able to kind of visualize how the ocean is evolving with time. So, yeah, those are that's the main, you know, I would say, competitor. The other main advantage that we bring, is that, as I mentioned, it's a three hour intervals. So we provide phishing hot spots, every three hours, whereas SiriusXM is twice weekly. So, ultimately, I think tools are useful, and there's no one tool that's gonna solve everybody's problems. And, again, I mentioned SiriusXM. I think they really are a pioneer in this space. And, we've just we've taken a pretty significant step forward, and, I think it's gonna be a great addition to the toolbox for offshore fishermen.
Tom Rowland: So this was your idea?
Taylor Shropshire: Yeah. So, the found science, I came on about three years ago, and they had a prototype called FishCast, that was really bare bones, but the but the concept was there. And then I, through our relationship with the ROS, was able to work very closely with them in saying, hey. I think we can make something really exciting here if we're able to use your historical phishing reports. And they've been great to partner with, and have huge respect in the industry. Yeah. So it brings a lot of credibility to the product. And, yeah, we were able to work closely with them and approach different, MFD companies, and Semrad was the one.
Tom Rowland: And so when you when you're at Fathom Science and you're there for I don't know how long you had been there before you started on this project, but is this something that you because you grew up fishing and and liked it, were were drawn to a little more so than some of the other things that they were working
Taylor Shropshire: on? Yeah. Definitely. I think, so as I mentioned, we come from, like, an academic background, a bunch of PhDs. It can get pretty nerdy pretty quickly. So, I mean, there's definitely people on our team who have not really fished before. So I'm, one of the the people on our team that brings that that experience and just a love of of the ocean, and ocean data. I've I've just been always fascinated by how complex the ocean is. People sometimes, you know, think about it like a big lake, but there's all these crazy currents moving around. Right? And there's tides happening. And the fact that we can build computer models that give us pretty good estimate about how the ocean is gonna be behaving today. This has just been something that fascinated me going back all the way into my undergrad, time.
Tom Rowland: Yeah. It's it's amazing. When you when you first start, you see this project. Do you have to kinda get some permission from from your your superiors to put all of your time or whatever time you you decided to align in this? So how did you sell that to somebody?
Taylor Shropshire: Yeah. I mean, our team is smaller. We are a tech tech startup from NC State. So, I'm very close with the main founders. I've worked with them on research projects for a long time. And, so when I was on the company, they said, I think, you know, this seems like a really interesting product that we'd come up with, and all the pieces are there. Right? It's the ROS data, our ocean forecasts. It's it's all just statistics. So for me, I saw it as this is just a tool that I think a lot of people would enjoy. And I like fishing, and I really like ocean data. So, that's how it started.
Tom Rowland: Did you have to convince them of the market and how many people might buy this product or or what Yeah. We did. Possible, ROI on this thing was?
Taylor Shropshire: Yeah.
Tom Rowland: Like So,
Taylor Shropshire: I mean, we know we did some revenue forecasting, and, that certainly went into it. When we look at just the value of oceanographic data, there's a pretty high market. Because like I said, the ocean, is a pretty data poor place. And especially people that are fishing in the top tournaments in the world, right, they're looking for any edge that they can get.
Tom Rowland: Any edge?
Taylor Shropshire: Yeah. I mean, small, small edge. And so it it it seemed like a no brainer, that this type of quality of fishing data could be extremely valuable.
Tom Rowland: Yeah. And people will pay for whatever. I mean, you know, like, if this makes it
Taylor Shropshire: I
Tom Rowland: don't know. As you're if you're a professional guide or you're a professional tournament angler or you're just a professional captain that's trying to make sure that when the boss actually does get on your boat, you are in the meet. Like, it's super important. That's job security. So it you know, it it I'm I'm assuming there's a subscription. That's what you
Taylor Shropshire: said. Right? So, yeah, there's a weekly subscription. There's a month.
Tom Rowland: You can just wait. You can just get it for a week.
Taylor Shropshire: Yeah. If you want. Yeah. Try it out.
Tom Rowland: So you're going to The Bahamas. You're thinking, I'm just gonna try this out for a week.
Taylor Shropshire: Yeah. Yeah. Wow. So it's a $100 for the week. It's a $150 for the month. That kind of scales, to be, you know, more, more less expensive as you go up to a year subscription.
Tom Rowland: What does a year subscription cost?
Taylor Shropshire: I think it's a little over, like, $1,200. So, but, yeah, it's, I think very competitive as far as pricing when you look at the fact that when you purchase a roster for right now, it's around, you know, $90 a day. Mhmm.
Tom Rowland: So you're
Taylor Shropshire: having, you know, somebody sit down and write out specific details about your fishing area, which is great and, is always gonna be the the highest quality. But if you wanna forecast and you wanna be able to watch the water every day, subscription for you.
Tom Rowland: Well, this is gonna be a big difference between when people are deciding about their electronics on their new boat. You know, do you do you go with a different electronics company or do you go with Simrad? Because, I mean, this is that's why they do exclusive deals like this. Exactly. On mapping, on different things, on on sonar technology, on all these things because that's really the differentiator. Yep. Sometimes people will choose one because it's easier to use, but most experienced fishermen are gonna pick their electronics based upon the the products that they can run on those electronics or or the ability to do what will help them in their own fishing. And not everybody fishes the same way. And like that, it's a good transition here because not everybody fishes offshore. Do you have any aspirations or any any kind of, even even see, opportunity in the inshore market?
Taylor Shropshire: Yeah. We're certainly looking into that. I mean, FishCast right now, it does have a little bit of functionality that's targeted to towards that customer base. So when you get on your MFD and you look at the, optimal catch location map and you look at the hot spots that our AI selects, we have three different filters. You can do a 30 miles offshore, 60 miles offshore, or 100 miles offshore. So you can basically, you know, select the the hotspots that you wanna look at. The data is really most appropriate for, you know, offshore, but analytics are analytics. And so if it's saying there's a higher likelihood of this being an, clinically good space for large game fish, it's probably gonna be a good space for other species too. Right? Because we're essentially trying to predict where baitfish are gonna be accumulating. That's kind of, one of the main features right now, that I guess would be slightly different from, let's say, SiriusXM, because they have species specific recommendations. Ours are based on just analytics of, okay, is this likely to essentially mold no bait fish? Well, I
Tom Rowland: mean, that's what that's what you'd mostly do when you're fishing. I mean, you see birds diving. Yep. You're not thinking that's gonna be exactly you're just thinking that's gonna be a good area. There's a lot of bait there. It's probably what I'm looking for is probably gonna be over there. And it gets narrowed down more and more depending on the area that you are. You're definitely like, well, that's definitely Mahi or those are definitely tuna over there. If you see birds diving in a certain area, a lot of people know pretty much what it is, but you're just trying to fish in a in a like, even when we're fishing on flats. Right? Like, you're fishing on the flats. If you get up on the flat and you're pulling around and you're not seeing anything, that's not as good as if you're seeing sharks and rays and and and birds and everything is on that flat. You're like, okay. Well, there's a good chance that there's gonna be permit here or tarpon here or bonefish here or whatever it is that you're looking for because it's a it's a lively, happy, flat. Like, everything wants
Taylor Shropshire: to be there. Yeah.
Tom Rowland: And and those things seem like they have the same kind of statistics involved in trying to find those areas, on any given day with with whatever the the weather, allows. But I think that there is probably I don't I don't know where the where the data would come from. I don't I don't know of a source of data that you're gonna find that's gonna be like what you found with Ross.
Taylor Shropshire: Yeah. No. I mean, that to find something identical, probably not. But ROS does do in short phishing reports as well. It's it's just not as big of a, I think as as as many of their regions that they focus on. So Right. There is a possibility to use some from rocks, but, the other information would have to come from, other data sources like fishery surveys,
Tom Rowland: and fishing guides. And that's when you
Taylor Shropshire: Fishing guides. Yeah.
Tom Rowland: That's when it gets really tough because as much as some fishing guides would like to to have this tool at their disposal, they're not willing to give up their fishing spots in there and and their historical data because that's what gives them the edge. I mean, it really does give them the edge. And and and guides that keep a journal are generally better than guides that don't. And at least in some way, shape, or form, whether it's just notes or like today, you can just take a picture. Right? You take a picture of your fish. The metadata on the picture tells you what time it was, what day it was, and you can look at it and you can you can get just visual cues of what's going on. The sky was clear, like, it really is a lot easier today even with just your phone. And you can put a voice note and you can attach different metadata to these pictures, and then you can I mean, it it really makes a a huge difference in trying to replicate that? I wonder, one of the things that I wanted to cover before we run out of time was I wanted to know what your AI journey was. When did you first discover AI? In what context? Machine learning, whatever. And, when where were you, in your scholastic career when when you first kind of understood what the potential here was?
Taylor Shropshire: So I did my PhD at Florida State University, ended up doing a postdoc, with Noah. And after that, got brought on as a research scientist at NC State University where the company was founded. And the research group there that I joined, they had started this is going back, like, probably 2020. Mhmm. So a little bit before the '8 you know, the AI craze really started. And one of the things about ocean models are they're very computationally expensive. So you can think about all the calculations that are having to be run. And, so we started thinking about, okay, how could AI be used to run these ocean models? So originally, these ocean models are run basically off of fluid dynamics. It's all physics. It's all, you know, with the temperature of the water, the density of the water, how the wind is blowing on the surface of the ocean. All of this kind of goes into it. So we took those historical models, and we used them as training datasets. So we basically said, okay. AI model. This is how the ocean works. Learn how it works. And then once it's done doing the training, we can now use it to do a a forecast. And So
Tom Rowland: so just just before we get ahead of ourselves, in 2020 Mhmm. That's what she said. Right? 2020. That's that's that's a long time ago in this in the in the world of AI. That's like that's like talking about, like, you know, a a long prehistoric times of of AI. So it wasn't chat g b t. No. Like, what what was an AI model, and what how was that even available to somebody like yourself in 2020?
Taylor Shropshire: Yeah. I mean, AI has been developing for many years. It just hasn't it just hasn't, you know, gotten the mainstream attention until something like Check GBT. And, AI is not just large language models. It's other it's a system of of models that basically can learn how patterns work. And it's you have to use a training data set that you say, okay, this is this is the correct answer. Right? Mhmm. And I want you to try to get to that same answer. Mhmm. So chat gbt was the part that really hit it off, and that's basically saying, okay. Somebody has a question. This finds similar, you know, text and answers to this question. What we were doing was using AI more on the data itself. So we're not asking it a question. We're saying, okay. I want you to provide a forecast. I want you to say, what do you think the ocean conditions are gonna look like in six hours from now based on all this historical information of how the ocean evolves through time. You know, the tides are pretty consistent. So that's kind of the premise. It's AI. It's just not the large language model part that you're familiar with. And the real advantage about bringing AI into the forecasting space is that once you've done the training, you can run those models in a matter of, like, a few seconds. So you can run them hundreds of times, which gives you, like, what's called an ensemble, which is a group of a forecast instead of just one. So you start to get some understanding about the probability of, okay, a 100 of the models said, you know, the current was gonna do this in six hours. That's gonna give you a lot more confidence than just running one kind of physics based model. I know that's a lot of technical description. It gets a little bit hard to follow, but, yeah, hopefully, that gives you some background.
Tom Rowland: No. I I it does. And then I'd like for you to just kind of stair step us through, like, from 2020 when you're doing that to, like, as as everything's getting better. Right? Like, your your models, your training your model to get better, but at the same time, you're having chat g p t and and other, kind of large language models out there that are also getting better. And I'm sure at some point, you're utilizing those capabilities with what you're working on.
Taylor Shropshire: Yeah. I think the when you're when you're thinking about kind of progression, one of the big changes that's happened is cloud computing. So used to to run these models, you have a big server at, like, a university or something. Now we just pay Jeff Bezos, and, we can have access to some of the strongest computers on Earth. And so we're able to run much more complex models. The cost to running those models is dropping. We can get in a whole discussion about, you know, data servers and the challenges that those bring. But the reality is that just computing has gotten more accessible. So that has allowed us to run more complex AI models and develop them.
Tom Rowland: So you also hear, about quantum supercomputers, and that's on the that's kind of the the next big frontier as far as I know. You're you're you're you're deeply in it much more than me, much more knowledgeable, but it does seem like quantum supercomputing is something that's real. It is attainable, and it is not that far off. How will all of our models, including, like, hurricane prediction models and everything, you know, being able to go catch fish, that's great. But I'd rather not
Taylor Shropshire: be in a category five. Yeah.
Tom Rowland: Right. Yeah. So how when we start to be able to access, quantum supercomputers, how will that change things in your opinion?
Taylor Shropshire: Well, first, I'd like to say this conversation is really interesting. In fact, we're going from fishing all the way
Tom Rowland: to quantum computing. Well, I mean, it, but it it it actually does have, you know, I mean, there it's it's the same conversation because like what what you're doing right now, if we were talking fifteen years ago, this would seem like something so far off.
Taylor Shropshire: And and
Tom Rowland: the ability to, like there have been major changes even for like an inshore fishermen to where I remember we used to we we used to, like, see a big storm coming, and we, honestly, we wouldn't know if that storm was a mile thick or 50 miles thick. And we would just have to turn around and go the other way, and we would just we were like, man, if we just knew, like, how fast that was moving or if it's if it's a big storm or if it's something to worry about, it would change our fishing completely. And, you know, not very long after that, you got you got it on your phone you can just look at it and go, oh, it's just a tiny little thing. We'll just bust right through it or we'll just sit right here next to this key and it'll pass right over and it'll be clear on the other side. Before that, we were running 50 miles in the other direction to get away from it. It. It changes everything. It changes safety. It changes the way you fish, the the amount of fish that you're gonna catch. Everything changes. And somebody that's really into fishing, like most of the people that are listening to this, advances so that they can go and catch more fish and do things that they want to do. Now, of course, we can talk about like the negatives to it, which there's positives and negatives to everything. But also, if you're really into fishing, a lot of times you live right next to the coast and you are in a hurricane zone. And the last few years in Florida has been really, really tough on a lot of people. But this year was quite a bit different. Mhmm. And it just seems like like this last hurricane that was coming into Jamaica, I mean, I'm looking at it. I'm going, dude, that is going straight for Florida. Yeah. And it goes straight over Jamaica, but they were so confident all the way. This is gonna it's gonna turn. It's gonna end up way out here. Yeah. And that has to be better computing and and the ability to run more data through through a system and get an answer very, very quickly.
Taylor Shropshire: Yeah. Absolutely. Absolutely. I mean, it's I like to tell people, imagine a world without, accurate weather forecasts. You know, you just
Tom Rowland: don't even have to imagine, man. There's there's a monument to it the Florida Keys where if you're going over Channel 5 Bridge, you look out there and there are these things that look like these giant big cement coffins, but that's where the old bridge was gonna come. And this hurricane was coming. They had zero forecasting except for somebody could be train and the train could come down there and literally one hundred and fifty people died. There's a monument to it out there, because all of a sudden the sky gets a little dark. You're like, boy, it looks like it's really gonna rain. And the next thing you know, you got a cat five hurricane on you in two hours.
Taylor Shropshire: Yeah. And Yeah. So that that advancement I mean, we're talking about catching fish, but saving lives. Right? I mean, that's it's it's not a hard jump. Understanding the weather, understanding the ocean is just a critical piece of information when you're making decisions on the water, when you're living near the water. And Cloud computing, AI, will only help to continue making those forecasts more accurate. So I can't speak to quantum computing side of things. That gets a little past, yeah, like, my full depth of knowledge. But, ultimately, you're hitting the nail on the head, which is that, you know, these forecasting systems will only ever, you know, continue to get better, and they're incredibly valuable for more than just fishing.
Tom Rowland: This year, they, and I've noticed this in a couple of different different aspects. Like, for one, I used to use this weather forecast called dark sky. Do you remember what that was? Dark sky would give you a forecast for the next fifteen minutes or the next hour, and it was unbelievably accurate at your location. It would tell you it's gonna stop raining in five minutes. And you're looking and you're like, there's no way it's stopping raining in five minutes, and then it would stop raining in five minutes. It was it was unbelievable. But for whatever reason, Apple bought that from whoever dark sky was, and they either killed it or something else happened. But there were several of my other weather apps that once offered, like, this short term forecasting, like, for the next hour. And then all of those all of that's gone away. And so this year, we were hearing that, there was gonna be some data that was normally available. And for whatever reason, it wasn't going to be available, and it was gonna change the the hurricane prediction models because this wet this this data that that anyone could access and bring it into the model was, for whatever reason, wasn't gonna be available. I don't know what I'm talking about. Do you know what I'm talking about?
Taylor Shropshire: Yeah. I think there were some changes, you know, some federal funding that there were some concerns around, the number of, like, observations that could be, done each day. I I'm not an expert in this area either. But, yeah, that's the idea is, mean, models are only as good as the data you provide them. And so you can have a great model, but if you don't have anything to feed to it, you can just really drops the value of it significantly. And less data means, you know, less accuracy. And, so I think there was some concern about some changes, to some federal funding that would impact how many observations could be taken.
Tom Rowland: Yeah. Well, for whatever reason, my apps aren't as good as they once were. And I think, you know, it's not the app's fault. I think it's that they're they're just not able to access that data anymore. And I just wonder if there's, something in your, like, you need to keep this Ross response or, partnership strong so that you can continue to utilize their data. But are are there any other things that could possibly be out of your control if
Taylor Shropshire: Yeah. I mean, we use, any public data that we can get our hands on. So buoyed observations, tide gauges. We use satellite imagery from NASA and other organizations. There is some European, satellite data that we also use, so it's not, like, just all US. But the short answer is, yeah. I mean, we're we're not paying for satellites to go up. We're, you know, modern creating all these buoys offshore. So that data, has to continue to exist, support not just what we do, but, obviously, all of the, other weather and ocean forecasting that the nation relies on.
Tom Rowland: Yeah. Well, I think what you're doing is really cool, and I think that, it's gonna benefit a lot of very serious fishermen. I'm not sure that that I see, especially at the cost, that it's that it's for the true beginner. Yeah. I think it might be over the head. I think maybe you need to be out there for a little bit and kind of understand the frustrations of not being able to find any fish. And then And then
Taylor Shropshire: turn to us. Yeah.
Tom Rowland: But the thing is is though that you can buy it just for a week. That's very unique. I haven't, I mean, I don't know of any none of my weather apps that we're talking about do that. You know, you
Taylor Shropshire: can get
Tom Rowland: it for a month or you can get it for a year. You can be billed monthly or you can do billed annually. But the fact that it's open that that you can do it for a week, I think, is gonna open it up to a lot of people trying it. And and trying it on SIMRAD, you know, that I think they're the best and they, you know, a lot of people have SIMRAD on the boat. So with that being said, is it available? Like, I know that these days, it's really, really important to update your software on your electronics, whether that's Lawrence or Simrad. I get a lot of emails about that all the time. Like, there's an update available. You need to update it when those become available so that everything that's already on it runs better, but then you can also run new stuff. How far back, will your product work on a SIMRAD device?
Taylor Shropshire: Yeah. I mean, I think it depends on the MFD. Like I said, it has to have the Neon software, which is fairly, new. But right now, you can go on the CMAP web site and order a subscription. You can either do it on your computer or you can be right on your MFD. It has that little C MAP app icon on your MFD. Just click it, log in, select your region out of the 133 that you wanna have subscription to, and then it downloads within, like, a a minute from, our servers. And now you have access to a three day forecast, temperature, ocean color, and this, the fishing analytics optimal catch location map.
Tom Rowland: And so that MFD is connected through Wi Fi, so you probably need to do that at the dock?
Taylor Shropshire: Yeah. So that's the idea. I mean, you can do it, have a hotspot to your phone or if if the dock if you're in a marina and has Wi Fi, It's a, I would say, a pretty big advantage to to this, product that you don't have to have offshore satellite connectivity. Just download all the data right before you leave the dock, and now you have to forecast it. It runs every day at around 2AM in the morning on our servers, and then it's ready in within about an hour.
Tom Rowland: Three days worth of data there?
Taylor Shropshire: Yeah. Three day forecast from when you download it. So you can it's there's five specific times, the 6AM, 9AM, noon, 3PM, and 6PM every day for the next three days. You can you can look
Tom Rowland: at. And so you said the CMAP website, but, yeah, I'm sure you could see all this on the SIMRAD website as well. Right?
Taylor Shropshire: Yeah. So Simrad is the soft I mean, it's the hardware. CMAP is kind of the software that the the charts are available on. But, yeah, if you go to Simrad, you'll see it. And they have a fish cast feature right now on their main page. Click on it, try it out for a week, and, I think you'll become a believer.
Tom Rowland: Man, that's awesome, man. Well, I really appreciate you, coming on and telling us the story. I know that a lot of people are gonna wonder what that is. And, we got a lot of boat shows coming up and SIMRAD will be at all of those boat shows, Miami Boat Show coming up in February. That's that's a big one, but I'm sure they go to many other ones. But you can go to the SIMRAD website and check it out. But thank you for coming on, Taylor. I really appreciate it.
Taylor Shropshire: Yeah, Tom. It's been a pleasure. I really enjoyed going all the way from fishing to quantum computing.
Tom Rowland: Well, it's kind of the same these days.
Taylor Shropshire: It's kind
Tom Rowland: of the same. All right. We'll be back with another great guest next week. So join us then. If you like this, share it with one of your buddies. Okay. We'll talk to you later. See you.
Subscribe to get the latest episodes, show notes, and exclusive content delivered straight to your inbox.