A Medical Record for Each Vine w/ Shawn DeMartino, Sentinel

Hey listeners, just want to let you know that we've released the latest compilation of show notes in book form, covering episodes from 2022 to 2023. It's full of insights on sustainability, marketing, and even has a few celebrity sightings. Pick what up on Amazon and support the show. Welcome to X-Shadow. The podcast that navigates the business of wine with unique perspectives and insights with your host Robert Vernick and Peter Young. Welcome to this episode of X-Shadow. Today we are starting our series on "Vitaculture and Winery Technology" and our first episode in this series is with Sean D. Martino, the CEO and founder of Sentinel, which is the first vine by vine vineyard management platform. Sean, welcome to the show. Can you please give me and Peter a brief overview of your background? Yeah, for sure. Thanks for having me first off. I've been in the wine industry now for 10 years. Really a good portion of that was in wine making. And then partway through I switched to "Vitaculture" and some research and development. And then, and they'll probably last third of my career, then more general management role. And along that time I've also developed Sentinel with my partner. And it was sort of a bit of a COVID project that kind of spun into something real, working in "Vitaculture" for a long time. I realized the need for a platform that a technology that could allow vineyard managers to collect vine specific information over time and not introduce sort of the onerous task of data management and an organization. So Sentinel was really a product of the last few years, built on, you know, almost 10 years of being in the wine industry, really on all sides, wine making, vineyards, and then now a general management role. So for our listeners who aren't familiar with Sentinel and you described it a little bit, can you just give them an overview of what the solution is and how is it work in terms of hardware, software, all that sort of stuff? And just kind of walk us through the solution? Definitely. Yeah, so Sentinel, I sort of, the analogy I like to draw, is Sentinel is like a patient medical system record, but for vines. So instead of human patients, we're talking about vines. And the reason why I like to frame it in that way is that when we think about vineyards, they're really a population of diverse vines, right? There's really, as time goes on, they become more heterogeneous. And so how do we easily collect and capture information knowing that over time these vines are going to have really interesting life histories that we want to capture? How we do that is through a mobile app and then a web-based desktop platform. So very simple. The additional component that became quite interesting from a technology perspective was the implementation of high accuracy GPS. And so we use these little receivers that clip to the back of your phone that augment your phone's geolocation. Why is that important? Because vineyards can be planted on very tight spacing, sometimes down to about three feet. Your phone's local GPS is accurate to multiple meters. So we don't have the resolution there to be able to interact with a very tight spacing of tight grid of vines. And so this, what's called our GK technology, uses array of satellites to augment that geolocation that's on your phone. And so we use that to know specifically where you are. And so we get down to sub centimeter levels of accuracy that allow us to map out all the vines in a field. And then also once those vines are mapped when we walk up to any vine in that fields, we recognize that you're at that particular vine. And we can populate that vine's record data collection forms that are all user configurable. Right. So we can really get down to that vine level. And then we consume out on a day, week, month, year over your basis to then analyze that that vine and the population of vines around it. So it really, like I said, kind of becomes this almost population health kind of problem or way of thinking about things, whether it's virus or production over time, vigor, etc. Where the insights that we're driving are based on individual levels of information at the vine level. But at the end of the day, because we're bit of culture, us, we're vineyard managers or wine makers, we're interested in sort of how these cohorts sort of assemble themselves. So it's a spatial component and a temporal component is kind of the broad swath of how we think about things. So you have someone in the field with their phone with this special device. And so it's populating all the vines. How do you get the data around the vines into the platform and what data is it collecting? Yeah. Good question. So the way that we are able to add data to the platform first off is establishing the base map of a vineyard. And so there are actually a not trivial number of clients in the world who have already sort of mapped their vineyards on some level. They have sort of the coordinate data of their vines and of their vineyards, particularly vineyards that have been planted more recently in certain regions around the world. We said that they already kind of have these coordinate systems for their vineyards. We just then make them more accurate by then shifting them into space where the actual vines are. We're also really interested in the elevations, not just the lat log, but also the elevation of vines. So we build a 3D model. So sometimes we're using that information that folks already have. A lot of the time that's not the case because we work now over five countries around the world. And so where everybody's at is very different. For vineyards that have never done any mapping before, what we do is we enable them to capture those coordinates for each vine point by two different ways. One is sort of our rapid mapping feature where we're able to map the end of the vines in each row and then leverage what we know the spacing to be of those vines. Read in sort of the elevation change. And that's how we're able to then automatically create that 3D model using those end points. For certain vineyards that are sort of terrorist or we have a lot of clients that their vineyards are planted 60 years ago. They're not on a grid. And it's all kinds of stuff. I mean, there is no uniformity of spacing. You have some vines or four feet apart. Some vines are six feet apart. Some vines are two feet apart. In those instances, that's where they'll go out and they'll manually drop those vine points. And so we've built spent a lot of time developing the app to allow for ease and speed of that initial mapping pass in the data collection. And so it's somebody with their phone collecting that information as quickly as they can walk that was really spent a lot of time on the hardware integration and the software integration to really enable that speed of data collection. And so that first pass is sort of a manual pass and then the data that you're able to collect on those vines going forward is everything from images to comments to discrete statuses that folks have control over sort of infinite control out of the box. Generally, everybody's interested in understanding production status. So is a vine producing or non producing? Is it a root stock? Is it a miss? Right. What's the light stage of that vine? Very important. The other key pieces, the virus and maladies. That's becoming more and more of an interesting challenge that folks are trying to wrap their heads around whether that be red blotch, pierces, disease, trunk disease, various things. So there's the maladies that we can capture more information on those over time as discrete statuses but also image data, et cetera. And all of that's fed into the vine record. So at the end of the day, while what you see is a dot map, colorize dot map of your vineyard, which a lot of folks are very familiar with, the end of the day what's sentinel actually is it's a record of each of those vines. So each vine actually has a Facebook page, honestly, where we're able to collect hundreds or thousands of images on every vine, as well as all of this status information that you can analyze over time and then visualize that as a map is charts, et cetera, to get to discrete counts. But the end of the day, the real interesting feature is that we're collecting this information on each vine's record. Each vine has its own ID. It has a location in space in the latitude, longitude, and elevation. And then there's all kinds of other parameters, the year that that vine was planted, variety, clone, all those kind of basic things. But then we enable you to add a million infinite amount of data on each of those vines, which from a back in technical perspective was quite interesting and challenging. Because as you can imagine these historical data sets, we anticipate then getting up to 75 years, 100 years of information right on some of our clients have well over a million vines. And so the scalability of the product became something that we were really interested in early days. And so who populates all that data do they need to hire you to do that? Or is it the vineyard management team itself that does that? It's the vineyard management team. Yeah. And so because we're operating on mobile phones, anybody in an organization can collect that data. And the real goal is Sentinel is to kind of meet people where they're at, right? And to try to democratize the collection of this information. There's a lot of interest, obviously today in the world around automation. And we, Sentinel are keenly interested in the space of automation around vineyards. The one thing that we recognize working in five countries, however, is that every vineyard is different. There are a variety of things that sort of challenge stream line solutions coming into the wine space. One is that there's just such a heterogeneous and diverse array of wine growers. Every vineyard is different. Every geography is different. The ability to pay in each region is different. And so with Sentinel, what we really wanted to do was sort of develop a technology on a mobile app with a small piece of hardware receiver that clicks to the back of it that allows everybody in a field who is already walking these vineyards if not every day every week to easily sort of collect this information as they're seeing it in real time. And like I said, everybody likes to collect different kinds of information. They like to parse that information differently. And so we're leveraging the humans that are already walking in the field to collect this data over the course of an entire year. What we really tried to work on was allow those users to in that moment walk up to that vine, not know, you don't have to know, oh, am I in row 52 or row 53 because I'm writing on a piece of paper or am I buying 300,000? 67 in this row or by and 368, right? By just walking up to that vine, it kind of becomes like, I mean, for gamers out there, it's like a Pokemon GO situation. You walk up to the vine, right? It's kind of an interesting interactive thing. - Pikachu jumps out and electrocutes you. - 100%. Yeah, yeah. You know, you're able to collect that data then in that moment and then that becomes sort of a powerful thing that compounds over time. But the key thing is that that data, everybody likes historical data until you're the person responsible for managing these data sets, right? Now you've got this Excel table that's, you know, 8,000 rows and was corrupted four years ago when a new intern came on, right? Like, I mean, that's literally the day to day of any advanced viticulture department. We wanted to take all that away because the irony is the data, historical data becomes more valuable, right? The deeper it goes, but at the same time, the sort of the tech burden of managing the larger and larger data sets also grows. So you have this inverse problem of while the information is becoming more valuable. It's also imposing an enormous cost on people who at the end of the day, like we're viticulturists. I actually hate the word data. I really hate data because I had the lived experience of starting out being really interested in vines and then you end up in a situation where you spend four out of five days in an office on Excel or managing some data platform that's keeping you out of the vineyards. We really tried to, like I said, sort of meet people where they're at, which is like already in the field with a mobile phone. And so people who just sign up for your solution, like how long does it take for them to set it up and get all that data in? Is it a big lift for viticulturists to put in an acre of vines or is it half an hour? - It's really quick. Yeah, so now a day is, like I said, we were really lucky at the age of the internet. We had our first international client in Mexico two weeks after the first article was posted about Sentinel. And so, in my partner, Christian, he speaks 11 languages. So he was very early on the train of wanting to internationalize the product. So that implementation is something we focused a lot on. So usually what implementation looks like is the first day, we sort of sit down with the team, we run through the product, we have training videos, et cetera. And then we work with them to input all of their block information. So what vineyards you have, what blocks you have, and then those discrete parameters, what are the clone rootstock combinations? Right, all kind of the bulk metadata. And then the initial mapping pass that a lot of people do is usually, like, just, I'll give you a use case. The last week, a client with about 100 acres, one in turn, map to 100 acres, establish the entire base map that will be forever at that site until they replant in one week. So one person, one week, to map 100 acres and put all that base information in. So all of those vine coordinates, we're talking hundreds of thousands of vines, is in the system as well as all of the default statuses that vineyard will start out as. These vines in this block are producing. These vines, this is a rootstock block, et cetera. And then over time is where they'll then sort of collect more information as the heterogeneity of those blocks sort of unravels. And we can collect that information via, sort of, like I'd said, you know, discreetly going out and collecting that information on a vine by vine level. But we also built in work orders capability. The key is really to operationalize data. Like data is just data until it becomes something you can use. And now it's information that's valuable. It's like really important to us. So the work orders component, the reason why that's interesting is that we're writing this information down to the vine level and irrigation conducted on an entire block on the primary irrigation set. Each vine that's on the primary irrigation set in that block will then have a record written to it. That on this day, this person, you're getting this block, you're getting this vine for gallons, right? But other statuses, right? Other operations like roging, planting, grafting, those inherently have the ability to update a vine status and important vine status. Let's say production status, the life history of a vine, a planting work order will take that vine position from being a mists to now it's a root stock or it's a bench graft, right? By conducting that operation by completing that work order, we'll automatically change the status of that position from being a mists now to root stock so that you don't have to go out there after somebody plants that position and then manually update. So the over time, the operations of the vineyard will automatically update these statuses, update the information on these vineyards and that's what becomes really interesting over time is that because we're collecting this information on these vines, ideally, we're targeting our operations towards those vines and then by completing those work orders we're automatically doing those updates. So it's the kind of thing that as it rolls as the heterogeneity of the vineyard sort of develops but as our data sets also following suit with that, right? It's a digital twin of the vineyard, ideally operations and work orders that we're using which at this point, a lot of folks in the wine industry have implemented winery management software, vineyard management software, I say the development in that space has been slower. Winery management software is I think the Christina Greater proliferation there but generally everybody's quite interested today in understanding sort of operations data and so we included that in with Sentinel because it's important but it also is something that inherently is an important aspect of vine's life history. What are the practices that we're performing in that vineyard and how are they positively or sometimes negatively impacting the health and productivity of those vines over time? - So in terms of the solution, I guess maybe we could do a comparison of like what are the core benefits of the solution? Say you were a typical winery that was just looking at large blocks in their vineyard. What does the solution unlock for them in terms of their vineyard management? - Yeah, for sure. So I'd say core benefits and I'd say the one that's really most motivating growers in the world today is an understanding appreciation and fear of the growing prevalence of diseases and viruses. It's a debate as to whether there's just as many now as there have been in the past where it's just identifying them as being problems now. I mean, you talk to some older growers and they'll tell you that the vineyards have always had probably some virus but we just didn't call it that. So on some of my wish we still lived in those days. Whereas now, whether it's red blood or pierced disease or variety of these challenges, growers large and small at the all ends of the quality spectrum are faced with this conundrum that it's very expensive to plant a vineyard. We do so with the expectation that those vines will be in the ground for at least 20 years. But I think even nowadays from a quality perspective, and an assisting ability perspective, you see the vast array of the wine industry is now really interested in old vines, right? And we see that gosh, we're in the cycle of replanting. Most vineyards every 20 to 25 years, but that's so interesting to make wine from vines that are 75 years old. I mean, the reaches of the world were over 100 years old in terms of the vines and those plantings. And so I think for the large grower, the grower who's more interested in sort of the scalability of farming, they're still actually trying to wrestle with these vine by vine challenges, right? Virus disease is a vine level problem. And so how are they able to over broad scale be able to capture this vine with virus and then we rogged it, right? See the spatial spread of these challenges over time. A tool like Sentinel allows them to do that to your point, the scalability. Are they going to be collecting two dozen different parameters on every single vine in their vineyard on day one, probably not, but they'll focus on the really important ones that impact the bottom line, right? The longevity of that planting is very important, as well as its quality too. But I think for the largest growers, they're still quite interested in and seeing the spread of these problems that, you know, even the smallest growers have the same issue. You don't necessarily notice a problem until it becomes big enough for you to notice. And that's an issue of the specificity, the granularity of the information that we're capturing is too coarse that it's not picking up these things early enough. And what we see in particular with viruses is it doesn't take much for that problem to grow from something that's very small and hard to see to being something that's economically a problem. So from a virus perspective, I'd say, it's the existential thing. So it allows you to see the problem and probably it's highest contrast. But at the end of the day, there are all these other really interesting things that are maybe less existential, but impact growers that are both large and small. - And so is the end goal then that the holistic vineyard has a longer longevity? Is that the be-all and all that you identify? Problems early pull out the disease finds replace it with a younger one and kind of preserving the average life of the larger vineyard. - I would say so yesterday I was taught to grow over 2,000 acres. And they're interested. The conversation, you know, if you didn't know going into the conversation that this person is over 2,000 acres, you would have thought they were a 20-acre vineyard. I mean, for these same problems of how do we extend? I've got my, it's always the everybody has the same story. What are my best blocks? Generally, they're on the best soils, but a lot of the time they're also, they happen to be the oldest blocks, right? And oldest can be everywhere from 80 years to 20 years. Sometimes even 15 years, right? And you say to yourself, it would be really bad if I'd rip out this block because it's the core of my blend. Or I really just from a financial perspective need to get five more years out of this block, right? And then of course you always have that. I just planted this block five years ago and I'm seeing this decline. I'm seeing a viral issue or whatever it is, right? It's kind of this issue that unites everybody is disease and virus and this question of how do we extend vineyard longevity, particularly nowadays, where the cost of development, of redevelopment is not going down, right? and the downtime associated with it. with redevelopment has always been a problem. And so I think it's the one kind of core issue that like I said, growers that are 10, 100 times larger than the smallest growers are interested in the same problem of either from a quality perspective or just from a financial perspective. I really, really want to need to be extending the longevity of my vendors. What would it mean if I could extend the average duration of a reply from every 20 years, to every 30 years from a quality perspective, but also from a bottom line perspective, that's very attractive. - Yeah, I'm just curious on how people, 'cause that everything costs money, right? So when people think about the return on investment, when they're doing, when they're gonna add this extra data to get that granularity that you're talking about, are they looking for improved grip quality, increased yield, reducing of cost, all of the above? Like what exactly, how are they calculating what they're gonna return on this investment? Yeah, how they quantify it? - A couple of different ways. Certainly, like I said, the one that's the clearest is when we're talking about the incidence of virus disease, roguing, right? A lot of folks are already pulling out vines because they're evaluating visual symptoms of virus or they're testing, which is already a very expensive process of manually going out digging these vines out, testing those vines, flagging those vines, right? There's already a lot of operating costs associated with this operation of trying to manage disease, which as we oftentimes see, they're spending the money, so they don't have the information to really sort of target that investment, and so from day one, sent to most coming in and helping them be more efficient with that practice. It's also helping them get ahead of the problem, because in the absence of this really high quality information, what we oftentimes see is you're spending all of the money to go out and manage this challenge, but you're accidentally missing some vines, or I mean, the one thing that we've definitely seen these days, which really complicates the problem, is that a lot of these viruses can be asymptomatic, right? So for at least the first year or two, a vine can be infected with a virus, but not show visual symptoms, no symptoms at all, right? For some duration of time. And so if we're not able to evaluate here as a hotspot, and then maybe let's test just outside of that hotspot to sort of understand what the rate of asymptomatic infection is, we could be doing all the energy and work of roging vines, but then actually missing a significant enough percentage that, and we see this a lot, actually, folks have been for five years, trying to manage red blotches, an example. They're roving vines every single year, but the rate of spread hasn't changed. What we find when we work with them is it's because, sometimes it's as much as 10% of the vines that are left out in the field are actually positive for the virus, they just weren't showing any visual symptoms. And so with Sentinel, we're able to sort of help them get ahead of that problem. We have a number of clients that we've been working with now for three years, and for the first time, they've actually gotten ahead of that challenge, right? And the reason why this is the most interesting thing to talk about is because you see it in real time, right? Be killing a vines, the spreading of virus, the decline of vineyards. But there's a lot of other quality related parameters that winemakers, vineyard managers too, but winemakers care a lot about, which is generally speaking, folks are picking vineyards based on bigger differences, based on flavor differences with Sentinel, we're able to capture that information and allow them to create these spatial maps of different pick areas that correlate to lots that get fermented separately, and generally are also associated with different wines that end up getting made and blended. And so Sentinel really becomes that first sort of starting point of understanding how those lots are coming together from the vineyard. - So you mentioned a few diseases already. Are there certain diseases that are not helpful or not able to track around value into the platform that you're working on, or is it just whatever data that they're interested in collecting? - It can be stored. - Any and all data, which like I said, was technically a challenge to build, but it was important because we, like I said, see it all the time, every vineyard manager, usually the vineyard manager on the winemaker have different opinions on how they want to track a problem. And as you learn more, you adapt to how you're managing a challenge, and so users have full control over the kinds of information they collect in Sentinel and how they collect that information. As an example, with a lot of these viruses or diseases, generally speaking, folks start out with a binary, something either as positive or negative, right? Or a virus or a disease in question. Over time, as like I said, we've learned things can be positive or virus, but not show symptoms. So now that's a third status. A fourth status can be that something was maybe tested and it was a weak positive, right? It might be the case that it was a false positive, right? And so over time, that information can kind of grow with your understanding of the problem. It's not like you're being sort of boxed into an ontology that's set at the start of the platform, right? That was really important because like I said, I see this all the time in the wine space. Folks tried to develop platforms, technology, solutions, but they try to box people in. And then ultimately, it ends up being this thing that maybe gets implemented for one year. And then it kind of gets used less and less until the point that maybe five years later, it's kind of an obsolete solution. And then folks move on. - And so are the use cases the same for high-end vineyards like that go into expensive wines versus more mass-market vineyards? - They can be. Yeah, like I said, I think disease gets a lot, disease, virus, maladies, you know, gets a lot of time and attention everywhere. I'd say that's the great unifier. That said, I'd say generally speaking when we're looking at sort of some of these smaller or quality focus producers or producers where maybe the white maker is walking the vineyard every week. And that's where we get into things like folks are mapping out different flavor profiles. So as white makers are going out and tasting, they're mapping out within a block, let's say this is where I have more of these flavors versus these flavors. And they're using that as a way to earmark the spatial variability of flavor development, of maturity development within a vineyard. And then they're using those to then inform how they're making their pick areas, their different harvest decisions. And so that's probably like a really niche use case. That's definitely at the smaller end of the spectrum, whereas larger growers are generally interested in kind of the more either existential questions or the things that are more generalized to a larger population of vines. - And you said you've gone global now to five different countries. Do the benefits of this technology differ geographically? - Like I said, the one thing that's so fun about the white industry is you get on a call and you start out, you look at these vineyards and certainly when folks start mapping, I mean, vineyards are planted and all kinds of different crazy places and ways. And so there's a lot that can surprise you on the surface when you actually get down into it. We're all wrestling with the same problems. Like I sometimes, it's quite interesting like with vineyards, we talk about them on the block level, which is kind of like very similar to how you would talk about corn or an annual crop, right? Like we think about things on a parcel level. And with vineyards, it's so not an appropriate way of thinking about it. And the average vineyard manager and wine maker is not, right? We think about them on a by-by-by-vine basis, but it's how do you scale that, right? Like we're talking about hundreds of thousands of vines in a lot of cases. And so Sentinel and technologies like this that I see kind of continue to get developed are I think really for the first time helping people interact with that diverse population of vines on a very intimate level, but then not get so sort of bogged down in the management of that problem, right? Like if you can't offer operationalized the information, then it's not worth collecting. And so that's the critical piece. - And do you bring in any macro data like whether if there's heat spikes or fires even nearby or things like that that maybe get inputted at a macro level across the vineyards or is it just that not part of it? - We've yet to integrate sort of like climate or a lot of that kind of information. 'Cause on some level, we sometimes struggle with this figure and how best do you want to write that to a vine level and what is that impacting? So we think, 'cause I mean candidly, we work very closely with a lot of our users. And as you can imagine, a lot of the early folks who found us were very specialized users. And so they think about these things a lot, certainly soil differences, et cetera. And so we've done a lot of custom development to prove out some of these early kind of concepts of, okay, of course, we can include any information and any and all information because at the end of the day, it's a large database that has a spatial component to it, right? There are a lot of kinds of spatial data that we can collect, but at the end of it, what we're really looking at is for the average person who's gonna come and use this tool, this feature, how are they operationalizing the information? For me, I sometimes I worry with time that we as an industry, and I see this a lot, we prioritize a quantity of data first, which is generally where you start. If you don't understand a problem at all, you're just interested in collecting as much information as possible, but you can sometimes drown out the signal when you have too much of that sort of noisy information coming at you. So for me, and I think part of this is like my partner and I, we developed a platform ourselves, like we wrote every line of code, and I truly developed the product for myself as somebody who'd wrestled with these problems for years, and so for me, I'm always trying to be very sensitive about not losing the thread, right? Like as somebody who still to the state manages vineyards, like what are the things that I really need help with, and where am I at right now with my labor demands, with cost of managing vineyards, and how can I really help myself to make better decisions, to be more efficient. So like I said, we're always kind of toying with different things, but in terms of what we actually, of operationalizing on the platform is always a function of, do we see the need there with users and with our own lived experience of managing a vineyard? Because like I said, pretty much all of our users are people who are sure they're at their computers looking at this information, but two seconds later, they're out in the field wrestling with a very practical challenge like this irrigation manifold just broke or I need to train 10 new people on how to prune a vine or something. So we try to be really sensitive about that. Sean, obviously our listeners are often in the trade and I'm sure that some number of them will be curious if you could talk a little bit about how the pricing works for this kind of solution, like what is kind of a recurring versus upfront cost, how does it handle versus different geography or different size of vineyards, how should they be thinking about that? For sure. So I mean, as a predominantly software-based solution, I mean we have a small harbor component, which gets FedEx all over the world. So on some level, the harbor is almost negligible. We're almost a pure software solution. So we internationalized real quick, within the first few weeks of an article being published on us. So we early on were very cognizant of the differences of geography is around the ability to pay, et cetera. I mean, the wine industry, like I said, very similar challenges, but wherever region is at, is quite different in terms of what the economics are. And so pricing, the way that we try to, our ultimate goal is to have this be really easy. So our pricing we base off of acreage, it's not based on users. Because like I said, what we actually want to do is democratize the platform. We want to, every vineyard hand and vineyard manager in an organization can have access to the platform and should have access and be collecting this data. So it's all based off of acreage, and that acreage price is quite small. I mean, it's usually less than a percent of 1% to the point of what the farm and cost for acreage. So it's very small. I mean, we're talking for even some of the biggest growers. It's maybe a couple thousand dollars a year. So it's a subscription fee and annual fee that doesn't change based on usage of number of users or the amount of data or images you're collecting. Because ultimately, like we felt that it was our responsibility to handle all of the scalability challenges of an ever-ever growing data set and how we handle data warehousing, et cetera. The end of the day, what we really just want users to think about is just how they're going to use the platform this year. And so it's an annual subscription fee that's based on acreage. And we really tried to run the company quite lean so that we're not imposing these extreme costs on people. Because we really are interested in developing technology for countries like Armenia. And as you can imagine, Armenia, the wine industry, in Armenia is quite different from the wine industry in the Napa Valley in a lot of interesting ways. And I love the fact that we have pricing that can be agreeable enough to folks in Armenia as it is to growers in the Napa Valley. So that's how we think about it. We never want price to be a barrier to somebody using this technology. Because at the end of the day, I sort of feel like it's actually not rocket science. Like it's a mobile app in the age of mobile internet. The RTK, geospatial stuff is kind of quite interesting. And maybe a little bit more-- you can get some technology bonus points there. But we're not curing cancer here. And so we shouldn't be pricing it like we are, is my feeling. And I think you see that a lot in the wine industry, folks from other areas of big ag that have much larger budgets and all kinds of efficiencies of scale come in and they develop technologies that are probably not price competitively. Or they find later on that the industry just for a variety of reasons. These are small family owned wine res-- most of the time. And these new technologies need to fit within their existing operating budgets. And as we all know, it's not getting cheaper to make a bottle of wine these days. So-- Yeah. And so given the benefits and costs, what are your clients seeing in terms of ROI? Gosh, I'd say, like I said, it's always easiest to distill it to something that's very existential and lands kind of on the bottom line on a day by day basis. And I'd say one of them is probably virus in the sense that-- like I said, we have clients who've been working with us for three years. And they were roguing sometimes upwards of 1% of their vineyard every year with a growing viral problem, whether that be red-launch or piercess disease as an example. And they were targeting all the symptomatic vines that they could see. They were even testing outside of that. We have a lot of growers now that are doing in-house testing to try to drive down the costs of lab analysis. But they were still scratching their heads and realizing that they really weren't looking the problem. And then they implement something like Sentinel, which allows them to have a system of record. That's as sensitive, but scalable as everything else that they're doing in the vineyard. And instantly now, they can collect very quickly this information that informs decisions that get made that have real impact on the bottom line. And so for a lot of those vineyards now, they've been able to actually get ahead of the problem in a couple of those properties in the last year or so have only rogued a handful of vines, whereas before they are roguing hundreds of vines a year. And so, like I said, virus is an interesting one because from an owner perspective, I'd say that's always where the general managers and generally the owners get most excited about Sentinel when they see these naps. And they see you're over a year spread of virus and then how, with one change, with one piece of information, they were able to make a small modification in how they were attacking the problem. And in one year, they were able to save-- I mean, at some level, all it takes is for some of these geographies where bottle prices are where they are. All it takes is just a couple of vines. And when you think about the economic value of that vine, over, let's say, a five-year period, if you have that vine in the field for another five years relative to if, for whatever reason, you weren't able to get ahead of the problem. You rog that vine. What we oftentimes see in a lot of these situations, an inter-planted young vine, isn't doesn't become a producing vine in the middle of that mature vineyard per 5, 6, 7. Sometimes those vines actually never become fully producing vines because they're competed by these mature vines that are surrounding them. And so a lot of times what we end up seeing when we come onto properties is some are upwards of 8% of these existing vineyards are either not producing or under-producing because there are these hidden issues of replants that were conducted 10 years ago that just never really established. I mean, that's so normal. And it's normal because how can you expect a grower to remember where the vines are that I planted five years ago in a 30-year-old block? What do I have to do? Walk every single row, how often every month to go check on those vines, give them a little extra love? That's unreasonable until you have a map on your phone that is a color-coded dot of where all those vines are. And then you can just send somebody out on an ATV to go do another scouting pass or something along those lines. I mean, this is where, like I said, Sentinel kind of grows and lives with people. On day one, are folks thinking about that? No. But by year two, year three, generally speaking, folks realize the power of the information that they have. That's just a digital representation of the vineyard that they have already that's able to surface to them these visuals that tell them something that they maybe already know, but just didn't really have access to before. So that's kind of what's quite interesting with Sentinel that we, like I said, we built it for longevity, which is something that the technology industry doesn't oftentimes think a lot about. It's always run, don't walk. And so you raise an enormous amount of money, you think about what the total addressable market is. And then you build a really aggressive funding strategy to take that market, right? And oftentimes what you see is these kind of boom bus cycles. And that's why a lot of startups don't necessarily survive is because you're not really thinking about sort of the longevity, right? You're in some level because of the way funding works. It's important that you're ever growing. And what we find with Sentinel that's quite interesting is it's word of mouth. And the industry is-- and I kind of like it, the industry is very skeptical of new technology. And for good reason, I think for so long, I would get technology thrown at my face. You get the sales emails and sales calls. You look at something. And after a 10 minute demo, you'd realize it's not actually attacking the problem that you need it to because somebody from outside the industry developed a solution that they cut a couple corners. Like I said, they tried to sort of build a fixed ontology that isn't sensitive enough to the problems that you're trying to tackle. So how are your customers hearing about you? Kind of marketing, are you doing to generate new clients? So organic search is huge, thanks to Google. I mean, that's where a lot of people find us. We've had some great articles written in 750, in Harper's Wine, podcasts, for sure. I mean, just organic search, like most of our users are online, and they're googling or searching for these kinds of technologies. That's huge. But then I'd say certainly within the last 18 months, like you kind of reach a critical mass of word of mouth. And so now I'd say that's probably our biggest source of new leads. And as a result, I mean, this isn't a joke. Our referral program is if you refer somebody to Sentinel and they end up signing, we send you a bottle of Krug because I mean, like, who doesn't like Krug? And on some level, it's like the best, in, I mean, most interesting way of rewarding that word of mouth referral program. And like, I've had that experience myself. If I see an ad for some technology, I generally won't take it seriously until I hear somebody else talk about it, right? It'll hear somebody else's used lived experience. And it doesn't take much. They could just say, oh, yeah, that thing's great. And then you sign up for a demo. the next day. And so not joking, that's literally our biggest source of new leads. So I hope our listeners, when they sign up, they refer, they use us as a reference. I never said to your 90s and I'm like, "Yeah, that's true." Please mention the podcast. That's a one for one. I mean, you guys could be swimming in that stuff. So, um, yeah. Daddy needs some crew. So you mentioned word of mouth. Are you publishing case studies or anything like that? Because I think some of the impactful like older clients that have been with you in multiple years, like actually getting down to the simplicity of like, how someone's used this and how they, what they were doing, what the cost was now, what they're doing, how they're, but they'll be a little bit more surgical with their roging of the binds. And so I'm just curious, like, what kind of case studies do you publish? Because I think you must have a ton of data that would be really interesting to other people. We have a lot of data, as you can imagine, one of the things though that we take very seriously is data privacy. And certainly around virus. I think it's changing, but the industry is still like, it's like nobody wants to admit that they have any virus at all. It's just like, you know, everything you're in this virus is part of living. And so that's one component from an ultimately, everybody owns their own data. And so that's been one thing that we haven't really focused on is publishing some of these case studies. Though we do have some clients that have a lot of insights that would be very interesting to share with the world. And so it's probably the thing that's most top of mind to me is somehow in the next year we could share some of those insights because they're actually quite profound. Like, I, there's a lot that's not obvious about these questions and challenges in terms of how do you control virus? And, you know, for real sometimes, you know, I almost caution people to take a slow approach. Certainly for folks who have been roging and roging and roging without information to really back up those decisions, you know, viruses taking down, you know, your vineyard, but you're killing it just as quickly. So I do feel that there's a lot of learning that has been achieved in parts of the world that's great to share. And so I'd love for that to be a part of the platform going forward. But certainly, you just, you know, you need to have users who are willing to share that information. Because ultimately, like I said, it's the data's theirs. It's not ours. So there are other vineyard management platforms out there. I'm curious on how does your product compare, or your solution greater to other solutions that are on the market? What are the key differentiators of why people should be using your solution? Yeah. I think the one thing that gets missed weirdly, even by people who are inventors is the importance of having a record of each individual vine, right? Having an understanding of a vine's life history, right? Sometimes people get really focused in on the end result, which is a, in this case, it's a dot map, right? You show somebody a map of a vineyard and where every vine is a dot, and it's the dots are different colors. And there are a lot of different ways to visualize information, but what the underlying data structure is, the underlying structure of information is what's the important thing. I can have a different dot map every single year, but if that's not tied, each one of those dots aren't actually tied to each other in an individual vine record that goes back is capable of going back into that entire vine's life history and forward, I don't really have anything terribly useful at all, right? Because at the end of the day, what I'm interested is understanding the longitudinal nature of these problems. I'm even your manager who manages vines that live for decades. I need information that spans decades. That is tied to individual vines. It's the same thing as if instead of healthcare being tied to your individual patient record, if instead we just made decisions on what Peter's healthcare should be based on just what, oh, the average person of X age of certain gender is. I'm not interested in what your sort of biometric parameters are. I just know that your X age, and this is that, and so, you know, no, I'm not going to see you this year, and oh, next year you should be on lipitor or whatever, right? That'd be a really bad way to manage healthcare, right? So I'm like my doctor. Yeah, exactly. I mean, I used to get your fluster out. Obviously, it's moving out there for a phone. Healthcare is actually managed. For sure. Unfortunately, at the end of the day, like it's actually, I mean, the reason why that's that's a challenge is because like, okay, you have all this information, but the kinds of decisions you make is actually like really nuanced and sensitive. And as a result, from a platform perspective, we want to give people all of the information that they find interesting and give that information time to expand and to build so that years later, they can look at it also with the lip history of that vineyard and make a decision about how I should be managing, let's say, Trump disease, right? Well, I know that this area of the vineyard over the last 20 years has always been weaker. That in itself will result in you making a different decision in terms of how to manage, let's say, the spread of Trump disease or utipa. If instead, that was a more vigorous part of the vineyard or part of the vineyard that was vigorous and now is weaker, right? These things can sort of like the thing that's quite interesting about vines, vineyards, and then certainly wine is that all of these things are so over determined. It's really important not to get so laser focused in on the one problem that's the most important thing to you today because very likely, the interesting result that you're seeing is actually the cause of many other things that have been happening over many years. And I spent many years of my career focused in on really discrete research and I was really trying to constantly split hairs and parse things apart and understand how each variable, right? From a first principles perspective, understand all these little things, time and time and time again, you'd spend two years on a project and realize that the native wisdom, the old farmers way of thinking about the thing was actually so close to the result that you achieved after 24 months of understanding every individual parameter that at the end of the day, like I really like to take a slower approach to managing vineyards and what I like about what we've built with Sentinel is that it allows you to do that. We're so focused on action. I identify a problem today. I want to act tomorrow, but these vines are really resilient and so as a result, sort of the strategies that implement in a vineyard should be tailored over that longer time scale, which means you need information systems that live with you. They live with the vineyard. They're infinitely flexible to the demands of what you're putting at them. So that's kind of how I think about it. It's an interesting space. My partner is like from the tech industry and so, it was I'm from the wine industry and so it over, you know, now it's been years that we've been working together. It's funny to kind of have those two ways of thinking about the world come together. He's constantly thinking about things in terms of these discrete systems and then for me, I'm like, well, it really depends and it will change. This is the right answer today, but I know in 12 months I'll learn something slightly new that causes me to think about it just ever so slightly different that now that feature is obsolete. So how do you build something that has like infinite flexibility in that regard and can grow with people as they're using the platform? So as you meet all these new vineyard managers and whimakers, what are some of the reasons they choose not to adopt your solution or what do some of the reasons they say, oh, I don't know about this and how do you address those barriers? I'd say one of them is is a question of time. I'd say and I feel this painfully these days. Labor is our biggest cost and ultimately too, even just as a vineyard manager, you have a limited amount of time and you have a way of thinking about things. I'd say the critical piece is actually oftentimes you have the experience of managing vines, managing vineyards where you have less control in a vineyard than you do in a winery in a lot of interesting ways. When you're sort of responsible for managing a vineyard, at the end of the day you always want to rely on sort of like your innate sensibility and a lot of times people are resistant to implementing a new technology, a new solution for two reasons. One, you're worried that it's going to introduce new demand on your time when you have already felt you never have enough time. The other thing is you worry that this new solution will actually give you less flexibility in understanding or managing your vineyard than you have today. That's real. The thing that technology is great at is it's great at systematizing things, but people hate systems, particularly in vineyard and farming space because you're like, I don't know, it depends. Yes, this is a block in a way. It's named a block, but actually it's these all these different zones. Then within those zones, these zones and those zones will change over time. I find the two things are it's a time and then it's an inflexibility. The inflexibility, like I said, that was early days with said to all. We sort of said that it's very important that folks are able to have basically infinite ability to customize everything that they're collecting in the platform. That honestly is really a non-issue. The time component is something that we've really tried to solve with developing really performance software that can scale. The goal there was, okay, these folks are, I don't want you to have to walk into a vineyard in an additional time because that is a time cost. How do I allow you to collect information when you're already in the field with your phone? That happens all the time. How do I allow data to be collected within the existing workflows of a vineyard using people? This is where the critical piece is that people can see and understand and recognize these problems, is that we don't have a way for them to record that information in a sensitive way and in a complex way as that information lives in the vineyard. That's what we tried to do with Sentinel. It was leveraging spatial data, temporal data, which is really what this problem is. It lives in time, but also in space. We're really just trying to bring that onto a phone, which like this is technology that's existed in everybody's pockets for the last 20 years. Not revolutionary in any way. And so, So like I said, as a technology, we're trying to just meet people where they already are. You're already putting flag-- if you're already-- I tell people all the time, if you have flagging tape in your vineyard, you're already doing vine by vine mapping. You spent all of the same time, if not more, time spending somebody out into the field to walk that row, to put a piece of flagging tape on that vine. It actually takes more time to tie flagging tape on the trunk of a vine than it does just walk up to it and click a button on your phone. It's the same amount of time, right? All we're doing is the flagging tape that color will change. The tape will fall off in a year or two with Sentinel, that information will live forever. You don't have to make a decision about that vine today. And on some level, you shouldn't. You can let that be something that lives. You can see how that problem changes and grows and expands. And then in 10 years, you can come back to that vine. That data is still there. That information still lives on that vine. And you could do something about it then. So that's kind of the way that I like to think about things. Part of my feeling with taking things slow is collect this information for a few years before you decide to go to the nth degree of the cutting edge of the technology that exists today. Everybody gets really excited about these buzzwords of technology. When in the end of the day, we're still understanding them. And a lot of these applications are incredibly niche, right? People need flexibility. Oftentimes, what happens when you run to the nth degree of a cutting edge technology as you find it, it frees you as much as it boxes you in, because you're kind of at the mercy of the development of that technology that's still being understand at a fundamental level. With Central, I think on what we tried to do is build something that's very well understood and can be extended to any application, because it's not something that's changing so rapidly that tomorrow the solution's going to look 10 times different than it does today. I think that's very important. These are slow, small organizations, and as a result, the technologies that we implement should be able to match them and beat them sort of where they are. Sean, I love that flagging tape example. It's a great example of like, hey, you're already doing this work. And our solution's just as fast if not faster. I'm curious, so since you've been doing this for a few years and have a number of clients with a quite a bit of acreage under what is next on your product roadmap? I'm assuming you're learning from your customers, hearing what they want, even doing some bespoke development for them. And then if it's good, you're going to kind of roll it back. So could you talk just a little bit about what's coming down the pipe for the Sentinel product roadmap in the next 12 months or so? Is there something that you can share? Yeah. The beauty and the curse of building the product, sort of yourself, is that you can do anything. But in terms of what actually ends up getting implemented platform-wide is a whole other thing. We've spent a lot of time on work orders to the point now where it's the kind of thing that we have users that have been working with us for three years, and they've yet to use all of the features there because we're constantly thinking about ways that operations in the vineyard will change information, fundamental information on that vine. And also, because we allow people to collect any and all information they want to, those use cases grow with the user base. And so we're always thinking about and looking at, what are the pieces of information that a user in Italy is collecting, and how is that something that's maybe a little bit ahead of the curve from somebody in the US or in Armenia, and then we're feeding that back and do, OK, here's a way that we should change the way that this operation is occurring so that it gives people even more flexibility. So I'd say the operations component, the work orders component is where I'm spending a lot of time because that, to me, is really the bread and butter of any vineyard management platform is how are you able to collect, operationalize and collect data on the things that people are already doing in the field. It oftentimes is the thing that we don't focus on. Generally, people think about it from a, I just need workforce tracking. How many hours did this person work in this field doing this activity? But it really just comes back to kind of a budgetary question. There's actually a lot of really interesting information in there for how we think about farming vineyards. And a lot of times, there's a lot of efficiencies that can be gained by looking at how and when we're spending our time in the fields, right? Efficiency is like an ugly word, but ultimately, what it really is is, could we be a little bit more specific about how we're interaction with the vines to achieve an even greater end result, right? To really target the vines that need the input the most today. And oftentimes, that's young vines that are interplanned in the middle of mature vineyards that if I don't interact with those vines this week, they will suffer some kind of micro injury that'll result in them not establishing to be a full producing vine in four years. And that's really important for the health and productivity of vineyards. So the work order, the way to think about it, is essentially the process in which the data that maybe starts off fairly homogeneous starts to diversify and kind of become very customized on a vine by vine basis in that work order in whatever shape or form is some kind of change to that data on a vine by vine basis. And then some kind of UI that's reflecting that change, so it's visually understood and groked by everybody who's using the app. And so you've put most of your effort in making that as robust as process, so it can handle many, many case studies and different parameters, right? And so that's the focus of that roadmap is making sure that that's leveraged down to that individual vine data set. - Exactly, yeah. - So you're seeing vineyards all over the world and wineries all over the world, and probably seeing a lot of other technological innovations from other companies. Are there any from other companies, whether it's in the vineyard or winery that you're personally excited about? - Gosh, I will say from a winery perspective, I give so much credit to wider-imagine platforms like in a vent who have been fighting. I mean, I've got to say, somebody who, I feel like even 10 years ago, I got into the wine industry a lot of people were using winery management platforms, but it was still so early days, which is crazy to say, 'cause I was like 2015, and in that time sense then, it's almost unthinkable to not have some kind of a robust barrel inventory system, right? Which one of the ways that I like to actually talk about Sentinel is it's kind of like a barrel management inventory, but instead of being barrels, it's vines, right? Like you'd be crazy not to know how many barrels you have, and which barrels are 10 years old versus two years old. So I've got to say the unsung hero of the industry in the last decade has probably been winery management software, is that have really revolutionized how wineries are being managed beyond that in my own day-to-day life. I really like soil moisture probes as a technology. There's been so much ink has been spilled on how to manage your aggregations and think about your aggregations, and it's a perfect example of something that we've been farming and giving water to plants for most of human history. And is it all that complicated? On some level, yeah, they're complex systems, but what are some simple ways that we can attack that problem? I really feel that understanding how different soil horizons are drying is something that the plants changing, but the soil structure within a parcel is not. And so if I can understand at this point of time in the season, given this a little rainfall or this irrigation depth, how quickly is soil drying down? Where is the moisture still in that field? Feels like a really important way to understand your aggregation is like kind of the 3D of where water is and not in a field. Feels like really interesting. I'd say that's something from like a technology perspective in the field that I'm really interested in, or just I use a lot today. And it kind of is another thing, like, not revolutionary technology. Let me tell you, it's been around for a long time, and it's telling us something pretty basic. And that's what I oftentimes like, is it's like, you know, it's basically something that is giving me information I can have if I just dug a pit, dug a hole, but like that's really time consuming. So instead, if I have a probe that's telling me how wet the soil is for a feed down, that's all I really want to know. Like if I could dig a backhoe pits every single day, I would do it. I really would. I mean, that's what I'm interested in. Where is water and where are my roots? That's literally what we're trying to understand. You know, and then there's the, what's the evaporative demand on the leaves, et cetera. Okay, there's all that stuff. Which once again, there's a lot of complexity there that we don't understand. Like there's so many variables that we sometimes lose the thread when we try to add, you know, thousands of variables into an equation where the weighting is something that we've just kind of approximated. Like, you know, sometimes it's like, you're doing surgery with chopsticks, but like oven mitts on. And so I try to eat sushi with chopsticks, but you have oven mitts on. That's like basically what farming is. I understand a really precise thing, but I'm limited by my inability to sort of like understand how it all fits together. The power of data to help make that picture come to fruition for everybody. Yeah, it's super detailed information. Sean, we really appreciate you covering in so much depth about what Sentinel is working on on a vine by vine basis. I'm sure it's very insightful for a lot of our listeners. But we'd like to wrap up every episode on a personal note. And we are curious is what is the most cherished bottle of wine in your cellar and when do you plan on drinking it? Well, it kind of ties two things together. It's like the most interesting bottle. I didn't actually have it when I had this experience, but it was something I found later on was it was actually, right when COVID broke out in Milan, myself and another friend were in Borolo, like that February of whatever that was at 2020. We were literally there, you know, visiting wineries. It was like an amazing trip, but literally it was like the day after we visited Contearno, COVID literally broke out in Milan and we drove back as quickly as we possibly could and got out of the country before we were like, are we going to get locked in here? So just wild point in time and it feels like a decade ago. But that's, yeah. trip, like visiting with him, it was like one of the most remarkable wine experience I think I've ever had. Like he's just an absolute genius. And talk about somebody who, from a technology perspective, and like integrating it into farming and wine making, like Roberto Pinterno, like with like his like automated tractor. And I mean, the guy is like a true Renaissance person who's able to like develop really complex these technologies. But then at the end of the day, make like really extraordinary wine and not get lost. Like it's so easy to get lost in like the minutia of this problem you're solving that you actually make really uninteresting wine. Or yeah, you don't integrate it together. And so like to this day, he's probably like the guru of like the way that I hope to like exist in the wine industry. And so after that trip, I of course like you know, bought a bunch of like Malforctino. I mean, they're really expensive obviously. So I was able to find a 1990, which is like not my birth year. Like we'll say that it is Malforctino that I have when I'll drink that, I don't know. I already got married. I don't know. Sometime in the future, I mean, those wines look forever. So I don't know how many kids yet. I don't know when my next big thing will be. But that's probably the cool one. When you sell your, when you sell Sentinel. Yeah, we'll see. I mean, yeah, we'll see 1990. I'll figure that one out. But that's like definitely the bottle that I'm probably, it'll outlive you. Don't worry. Yeah, probably. Probably. So that's, that's my answer. That's a good one, I don't know. But he's a legend. So. Yeah, that's great. Well, we want to thank you for having you on the show and sharing so much information about Sentinel and everything you guys are working on and what's coming down the pipe. We really appreciate it. Thank you so much. This has been really fun. [BLANK_AUDIO]