Tony Zayas 0:05
Hey everybody, welcome to the tech founders show, where we talk to people who are on the bleeding edge of technology, doing amazing things, changing the way we work, the way we live, the work we play, and more. So with that, Andy, how are you doing today?

Andy Halko 0:21
Today is a great Tuesday, another fantastic show lined up. So I’m excited. How are you?

Tony Zayas 0:28
I’m good. I just, you know, trying to figure out something. I was wondering if you’ve ever taught your computer how to taste wine because I need so I need a little help with that.

Andy Halko 0:40
No, I haven’t I haven`t. So I’m curious. Can that even happen?

Tony Zayas 0:45
Ah, well, you know what, according to today’s guest, it can so we have the right person to answer today’s challenge that I have. Katerina Axelsson, she’s the CEO and founder of TASTRY. And it’s the firm of the world’s first AI driven sensory scientist company. And I will let her come on and tell us what that means. Hey, Katerina, welcome.

Katerina Axelsson 1:12
Hi, thanks so much for having me.

Tony Zayas 1:14
Yeah, thank you for being here. We’re super excited. So I saw on your website, I thought it was pretty cool. It says that, you know, TASTRY is an innovative startup that is teaching computers how to taste wine. So tell us about pastry, tell us about the concept. And super excited to hear all about it.

Katerina Axelsson 1:33
Sure, um, I think I should start just by telling you the origin story and how this came about. And I thought it would make the most sense as to why we had to teach a computer how to taste. So I, I went to Cal Poly University, which is in the central coast of California, it’s an engineering school. And I paid my way through college by working as a quality control chemist in the wine industry, because well, we have 400 or so wineries in a 20 mile radius here. And you can say I was given a lot of freedom to run my own mad scientist type experiments in the lab in my free time. And I took full advantage of that. And I had come up with several inventions in the lab early on even before TASTRY. So I invented a contraction, for example, that used 35% Less sulfur and made a better quality wine and things like that. And TASTRY was one of those inventions. And it came out of a an observation I made about how idiosyncratic you could say the product development process is not just in the wine industry, but but that is where it started. So when one example is we had a like a half million dollar 100,000 gallon batch of wine, so so nothing romantic about it. And we sold half of that batch under one label, and then the other half of that batch under another wine label. So the same product went out into market with different labels, different pricing, different marketing, and the critics scored one of those wines, much higher than the other. And consequently, one of those products did really well and the other one didn’t. So there’s very much this fog of intuition, I think in for manufacturers and retailers and distributors on how products are going to be perceived in the market by consumers. And how do you, you know, how do you predict that? And that just led me down this journey for the next year and a half where I ended up talking to the head researchers of let’s just say major companies and flavor and fragrance companies and trying to understand how are they mitigating the risk of launching a product, that’s not going to work out in the market. And it’s a big problem, actually, because the current failure rate is about 85%. So I found that what they’re doing is is it’s not really working, it’s not really improving the odds. They’re looking at the chemistry of products, and they’re trying to identify the presence or absence of specific compounds. So for example, does benzaldehyde equal cherry and a red wine, and they just keep doing that over and over again. But the problem is, there’s a flavor matrix effect happening. So there’s hundreds of other compounds in this chemical soup at various concentrations and ratios. And it’s either masking or allowing the compound you’re looking for to be expressed. So it’s not very predictive. And there was no known method to look at all the chemistry in in the correct ratios the same way the human palette would. And then the second problem is we all have unique biology so even when you you know work with expert tasters, three of them can say something tastes like apple. Three of them will say it tastes like pear and there’s no ground truth. Then the third problem is where we really come in, which is, we found that even if you can predict what something is going to taste or smell like, it has very little impact on whether or not you actually liked the product. And at the end of the day, that’s what really matters, right? It’s not can I, you know, taste cherry or I don’t know, chocolate in my coffee, it’s Do I like it. So my methodology that I spent a year and a half developing was designed to break down the flavor matrix the same way the human palette would. And it was a very robust data. So I was looking at the chemistry down to like the molecular count of each compound. So after gathering this data, I realized it would take me like a year and a half to process like one component of the data set in like 1000s of years to process all of it. So I set up a meeting with the head of the master’s program in Mathematics and Computer Science at the University. And I just said, like, help me understand this. And he said, I’ll talk to you for 30 minutes. But after looking at the data turned into an all day meeting, and he cancelled his class and brought other PhDs into the room and long story short, Professor Dec. Tr, and I partnered up and what he did over the years was develop AI and machine learning algorithms, specifically designed to connect my flavor matrix data to consumer palates. So we combined our two inventions and filed for a patent. And we’re awarded that patent for all sensor based products. And that’s why we say we taught a computer how to taste.

Tony Zayas 6:42
That’s pretty amazing, awesome story.

Andy Halko 6:45
Yeah, we almost like kind of just flabbergasted at the story is, you know, one piece that I’m curious because it you know, you brush through quickly is real timeline. So, you know, when you how long did it really take to go from like, Oh, I’ve got this data to a real product.

Katerina Axelsson 7:05
I’m almost three years it was, was it wasn’t a business, the technology definitely came before the business. And it was just a research project with flavor chemists and sensory scientists and PhDs. And, yeah, we weren’t thinking about the business aspect of it for for a long time. So I think I incorporated the company in 2016, just to get a loan on my own analytical chemistry equipment. But we didn’t go into market until early 2019.

Andy Halko 7:38
I’m kind of curious about that, too. Is is were when you were originally thinking of this, was it to create a business and to do this, or was it purely scientific and you were interested in the topic? I’m just wondering what you know how early the intent was?

Katerina Axelsson 7:55
Yeah. So I have always been interested in business and starting a business, or joining and running a business. But I didn’t go into this wanting to start a business for the sake of starting a business. So it was it was quite academic in the beginning. And it just happened to be that the invention we discovered or stumbled upon had some amazing business application. So we just kind of took it and ran with it, if that kind of makes sense. So I guess, I guess, right. Person, at the right time. Had a role in that. Yeah.

Andy Halko 8:36
And I think that’s the point of kind of, kind of getting to a little bit is, you know, it’s interesting for folks that come up with a great technology that sometimes they’re not intending to reinvent an industry or anything like that. It’s just, you know, the the kizma that happens to do that.

Katerina Axelsson 8:55
That’s right.

Andy Halko 9:01
I’m also, you know, interested in this is, it’s complex. So what you’re doing is really complex. How did I don’t know how do you go about that process of really breaking this down into something that’s consumable? And I guess over a three year period steps, but, you know, how do you take something that to me as a as not familiar with the industry seems very daunting. Yeah. And how do you even attack that it started going after it?

Katerina Axelsson 9:33
Yeah, I think I think the amount of time it took played a role in that, um, I think anytime you develop any new method to look at the chemical properties of something, it takes a long time, it could take over a year to develop a new methodology for for anyone. I just happen to do it. Operating under assumptions that maybe weren’t so traditional. Let’s just say I took some creative liberties. And I would say that that took even more time, I think things really started to accelerate when I, you know, I brought the PhDs in, and they started to help me understand how to efficiently process this data. I mean, I, as I mentioned, just to understand the chemistry that was coming out of my method would have taken 1000s of years. So get getting someone who understands data science, how to process datasets, I think was really, really critical to moving that process along, figuring out the business model as well. And all the you can imagine there are a lot of applications and use cases, for this technology, figuring out which ones to go after and which ones are going to provide the most impact. also took time, I think we’re still in the process of figuring all of that out. While you know, focusing on on our current objectives.

Andy Halko 10:53
Yeah, I was gonna ask that is how applicable is to other, you know, foods and area? So what is the breath of this technology? You know, is it just moving to other liquids? Is it to solid foods? Is it way beyond that of something that I’m not even thinking about?

Katerina Axelsson 11:12
Sure. Yeah. So So great question. So we’ve evaluated and proven efficacy for all alcoholic beverages, aqueous solution. So I guess beverages of any kind. We’ve done fragrance, I guess, coffees included in that and hot sauce. We’re still researching, like more solid type food products. But if there’s any flavoring and like a CPG product, we can we can contribute to that whether or not the texture plays a important role, we still have to figure that out. But but from a business perspective, just the alcoholic beverage market and the amount of waste and going on here it like it’s a big enough opportunity to keep us busy for quite some time. Our technology applies to anything you can taste or smell. So I would say every couple of months, we’re evaluating what other verticals or products this applies to. And so far, we haven’t failed at any efficacy tests. So yeah, seems to be seems to be pretty broad.

Tony Zayas 12:23
Katerina, to this point, who is the target audience that you guys have gone after? Where’s that been focused?

Katerina Axelsson 12:29
Yeah. I’ll try to answer this in steps. Because it didn’t, it did happen in steps. So when we had this technology, the first thing we realized is we have this challenge and advantage, right. So the challenges are the advantages. We have this super high quality, robust, proprietary data set. And that’s really important in AI nowadays. And the challenges is we can’t buy this data, we can’t scrape it, we have to go out in the market and generate it ourselves. So the first product that we ever launched was a API and interface based personalized wine recommender engine. And it was to first evaluate the efficacy of the product in the market, but also to gather consumer pallet data in the market to get insight and validate the technology. So what we did was is, we used a different form of algorithm. And we we built it on top of our chemistry and palette data set that we had been training in house from that point, and we created a you can say, deceptively whimsical looking quiz. It takes about 20 seconds, it has 10 questions, and we ask customers things like how do you like black coffee or licorice or dark chocolate? And it’s dynamic? And it’s based on the inventory you’re looking at? So once you answer these 10 questions, we found we can understand about 85% of your palate the first time around, and in return you you get wine recommendations for any products you’re shopping for in the store. So and I can I can totally dive into exactly how that works. But the business benefit was obviously you know, we’re mitigating the cold start problem typical of most recommender systems, right? We didn’t want to wait until you know, we were a Netflix or Amazon where we had millions and millions of users and could collaboratively filter them so so that’s what the quiz did is it goddess data now, and we started with grocery retail. And immediately we were seeing really positive results. So there was already a use case for retailers because we were increasing wine sales. We were organically increasing their margin and customers were scoring the products on average 45% higher. So that’s that really validated the use case for you know, selling more wine and connecting consumers to products they love. It definitely evolved from there. We, today, you know, we’re vertically integrated business, and we work with distributors and manufacturers. And but we needed that data first.

Tony Zayas 15:14
I would love to hear, and I’m sure it’s very complex, and a lot goes into it. But just the high level about that quiz, and how you guys arrived with something so effective and so sure.

Katerina Axelsson 15:27
Yeah, um, so I have, I worked with a flavor chemists and people in sensory science, and we had a question bank have hundreds of possible flavors, and they’re acting as analogues to the chemistry of, say, the wine. So what’s happening is, is when you’re answering those questions, we’re plotting your unique palette in a multi dimensional space, along with the chemistry. And what the purpose of the machine learning algorithms is to do is to identify vectors in that multi dimensional space. So a vector can be a compound or a combination of compounds. So what it’s what it’s not doing is filtering by tasting notes. It’s actually looking at how you’re answering all those flavor questions in combination with each other. And then and then understanding how that relates to the flavor matrix. So yeah, it is a little complicated. And I can’t exactly tell you all the relationships other than the chemistry of a product and whether or not you like it is not a one to one relationship. And that’s the purpose of the machine learning algorithms is figure that out if I could comprehend like, if my puny human brain can comprehend that how the chemistry exactly relates to you liking or disliking certain flavors, I wouldn’t need AI or machine learning to do that. But that’s, that’s, that’s at a high level, how it works. And the benefit is, is worse, it’s so granular, because the chemistry is so robust that your palate is almost like you’re a fingerprint, we found so so we’ve identified a minimum of 119,000 Customer palettes to date, that are unique, and the largest co occurrence of people who have the same palette using our method is 13 people.

Tony Zayas 17:21
Fascinating.

Andy Halko 17:24
Yeah, that’s really cool. How do you see like that, and I know you’re not as much on that computer science AI side, but that technology, you know, evolving? I’ll be honest, you know, Netflix, and Spotify still don’t necessarily always give me the right, you know, movies or songs that I’m looking for, as recommendations. So how were How do you see that technology growing and impacting compared to where it is now?

Katerina Axelsson 17:53
Um, so I hope I’m answering your question. But on the next Netflix piece, we didn’t go that route, because we feel like what we’re doing is more accurate. Our algorithms or information based algorithms are more accurate than collaborative filtering algorithms. So if that’s what the Amazon and Netflix’s of the world do, and I would say, generally, it works quite well for you know, things like toaster ovens or whatever it is you buy on Amazon. But but when it comes to something like wine, they’re waiting variables in the data play a huge impact. So I can’t just look at someone who’s very similar to me, and, you know, has the same income and lives in the same zip code and is has the same demographics. And they, Oh, this person buys I don’t know, Range Rovers and Radiohead albums, therefore, I’m going to recommend this appliance to them. Because I can go out to dinner with that person. And we can share oaky buttery Chardonnay, and she could love it. And I would absolutely hate it. And there’s nothing in the data that could predict that. So I think the future in general, maybe even in AI is going to be for multidisciplinary companies where they’re gathering more robust data sets that will give you the ability to provide more accurate recommendations, more granular recommendations, and the more products we get out into the world and the more options we have, I think the more important that’s going to be.

Andy Halko 19:33
Yeah, it was a great answer. I think. I’m interested next in the entrepreneurial journey. I just was talking with someone the first business book I read was E Myth. And it was all about this concept that you know, you’re a bread baker and you you love baking bread, but then when you start a business, you’re doing HR and sales and all this other stuff. So for you that obviously came from the scientific basis background and a passion and interest in the industry. What’s been your entrepreneurial journey? And are you still involved in that? Or is your role and the way you you work completely different now?

Katerina Axelsson 20:13
Oh, completely different. It my role has changed. I’ve had several iterations for my role. So I started off as, as a chemist, but I only have a bachelor’s degree, right. So maybe in the beginning, that worked to my advantage, because I didn’t come in and try to invent something with preconceived notions. And I wasn’t told what I can’t do yet. Right. So that was good. But as things progressed, I needed more expertise. So I hired way more qualified and experienced chemists than myself to take over. And then they needed, you know, PhDs in computer science or AI. So we needed those experts. And as the team grew, I kept switching into roles where we needed help in the company. This is like really early stage. So for a long time, I was thinking, How is what I’m currently doing going to be a job for someone else someday. And eventually, I really streamline streamline my role to, I guess, be the person who’s getting all these brilliant minds much more brilliant than myself to collaborate on the same vision in the same room. So yeah, very different role than what I started with.

Tony Zayas 21:30
Katerina, what is the team look like now?

Katerina Axelsson 21:33
Yeah, um, we’re a team of 28. And we’re quite tech heavy right now. So we have a few salespeople and Customer success people to deal with our growing client base. But other than that, we’re, we’re chemists and engineers and data scientists.

Andy Halko 21:54
Now for for a group like that, how much have you taken into consideration like culture and building the team and creating a great environment? What do you do from that perspective?

Katerina Axelsson 22:08
I think culture is extremely important for a company, any company, let alone a company at this stage. So I would say we take company culture very seriously. The best way to have company good company culture to begin with, I think, is to hire people who who would fit in your company. That doesn’t mean that there isn’t diversity like that we have a very diverse team of people, but but we all work well together. But the one thing I think we have in common that we place a lot of importance on his being a self starter. So we were very fortunate to have people who are already motivated and believe in the vision. And I think the way we foster that is, in part is by culture, but it’s also the mission. I’ll give you an example. And I don’t think you can attract the most brilliant minds to work with you with I don’t know free food or cool gadgets in the office or anything like that. That’s definitely not what we do. We don’t bribe genius in that way. I think the people we have are standing behind what we’re doing, and they feel like they’re making an impact impact, and they’re making a difference. In the company, they have a huge role to play, if that kind of makes sense. So I think people are motivated by that by having that impact. So I hope I answered your question. Um, we’re also very resilient, I think, in our culture. So we’re communicating our wins and losses on a regular basis. And we’ve kind of built up that resilience as a team. We think that’s really important because well, I personally believe being in a startup is like, it’s like being in the Navy SEALs, at least psychologically, and you usually fail by quitting. And the way you don’t fail is by iterating, through whatever isn’t working. And I think that’s a huge part of our culture. Sorry, for the long winded answer.

Andy Halko 24:15
No, I think that’s, I mean, there’s so much to unpack there. I think I’m first curious, you know, how did you How do you develop that? Why and that vision and how do you communicate that people get excited because I agree with you, it’s not about the snacks. It’s about, like, why you’re there and what you can contribute to the world in a lot of cases. But how did you know make that tangible and communicate it to the team to get them excited?

Katerina Axelsson 24:45
Um, so we, we make it a point to communicate frequently, but efficiently, I would say so the more you’re in touch with each other, the more you get these big visions and ideas across we set time aside to get together and make sure we’re in alignment. That’s a basic one. But I hate I hate to almost say this, but like the I personally feel like the real key to having a good company culture is just to start out with great people in the first place. Everything else was kind of falls into place, if you have self motivated people who already believe in what you’re doing, like if you have that maintaining the cultures, I would say relatively easy. We’ll see what happens as the company scales, if we can keep that going. It’s working really well for us. Now I’ll give you an example on how I know this person is right for the job, if that kind of makes sense. So so we had a data science intern who worked for us for a year, when he was finishing his his degree. And when he when he graduated, he was getting ridiculous offers from all the top tech companies like, like, there was no way I could match that offer, or even come close to like half of it. And he approached us and asked us for a job offer. And I told him, I’ll try, we’ll, I’ll see what I can do. So he showed me his offers. And I came back with something that was like 40%, that was my best offer 40% of his lowest offer. And, and he came back and he said, Great, I accept. And I said, Great. We’ll get you set up. And then he came back and said, Okay, how much of my salary can I trade in for more equity in the company. And I was like, that’s the guy, I want to be part of the team. So when you find someone like that, you really want to get them in your team, because they’re going to be in it for the long haul.

Andy Halko 26:49
That’s a really cool story. I love that. Yeah. Go ahead, Tony.

Tony Zayas 26:55
I was just gonna ask along the hiring process, you know, when you’re talking to talent, what are some other things? Like, that’s a fantastic example. What are some other ways, some of the cues that you get you just, you know, people either get your vision and buy into it, or they don’t, I would love to hear just a little bit of that.

Katerina Axelsson 27:16
So there’s one thing I do when I’m interviewing someone, I try to convince them not to join the company. So this definitely worked in the earlier stages. Now, I think as we’re growing, we might change our strategy, but I would I would tell people, Look, are you sure you want to do this? You’re going to be working crazy hours. The pay isn’t that great right now. But you get equity. You know, there’s, you know, not going to be a lot of direction. Like there are a lot of things we’re all figuring out right now. Are you okay with that lack of structure? Like, are you sure you really want to do this. And so far, it’s worked because we have really good retention, and the people who hear that and lean in. First of all, they know what they’re dealing with. And second of all, I think they’re not surprising, they’re prepared to deal with it. So that’s one way we filter out. You can say startup, true startup people.

Andy Halko 28:21
You brought up the resilience thing. And you know, there’s always wins and losses. Right. And, but I think one of the challenges is for other founders, they hear about the wins more often. I don’t know, my I don’t know, if you mind sharing a story of like resilience, or, you know, what are some of the challenges that you ran into, you know, over this period of time, because I think that’s where people learn the most?

Katerina Axelsson 28:46
Hmm, yeah, um, well, so when I first started this endeavor, I didn’t, the original, I wouldn’t say founders, but people involved. It was just like a, like an extension of like a school activity or project. We were all you know, just getting together trying to figure out what to do with this. I didn’t realize how much you really have to filter for people who understand that you need to stick around for the long haul. And really invest, you know, sure you can you can, you can be part of a fun project for six months. But what happens when you graduate and your parents want you to get a real job or you know what I mean? Or when you realize you’re going to be working 14 hour days for the next couple of years with very little pay and what’s the opportunity cost of that right. Essentially, the original people who were part of this I don’t know what this project is phase, didn’t stick around and one of them was helping us in develop the analytical chemistry methodology that I started with further. And one day, he just opened a left and said, I’m sorry, I’m getting a lot of pressure, I have to go, you know, get a real job. So that was very difficult because we hadn’t, you know, created standard operating procedures are documented every step of the process yet. So we had, I had to go back, I hired more qualified people. And that really helped. But I had to go back and backtrack and figure out how to pick up where we left off. And that was a really a challenging time, but we came out with it with something better than we had originally. So there’s that. So yeah, I would say that one of the hardest things a founder can do is pick the right founders.

Andy Halko 30:54
Yeah, it’s pretty great. So what’s the I almost want to ask about the future and what you see for the company to potentially work backwards? But what do you see happening for the business in the next three to five years?

Katerina Axelsson 31:12
Yeah. So if you can give me a few minutes. So what what has happened and this will go into what is happening is, since we launched the recommender, we started generating this heat map across the US and now in Europe, on consumer pallets, right. So imagine, like comprising coverage map, but we can look at any product. And we can see what percentage of the population on a store local regional level this product matches to so as we evolved, we started to take that data and provide it to retailers who started to use it to optimize their assortment and reduce waste on the shelves. And then, toward the end of last year, we started to officially vertically integrate and work directly with manufacturers, it was opportune because COVID happened and E commerce exploded. And for many brands, this was the first time they were trying to figure out a new way of making money. So we took that same heat map we had generated from providing recommendations, and started to use it for identifying market of opportunity predictably, across the US, like we know your brand is going to succeed in this particular market. And you might be trying to sell it in Boston. But hey, at South Beach, Florida, you have 5x, the customer base, and your competitors are not very present there. And these are the type of people who would buy your product again. And then it just started to escalate into Wow, we can test the chemistry of a product and predict with 93% accuracy, how consumers will score it out of five stars, why don’t we do computational blending and formulate help the manufacturers formulate products so that they actually succeed in the market, and they’re saving millions of dollars of investment, or risk doing that. And so, at the end of this, where we are today is we’re working with the top six manufacturers in the US two out of the 10 largest in the world. And we’ll be moving into Europe, late q4 of this year, and now that we’re vertically integrated, and we’re helping people connect to their customers mitigate waste on the shelf and mitigate the risk of launching a failed product. Those are the three things we do, we’re starting to see that we’re generating this flywheel effect with the data where the more retailers we’re working with, the more valuable we are to manufacturers, the more valuable we are to retailers. And we’re in a sense starting to connect the two in the supply chain in way in a way that has never been done before. So I think the next step is to move into Europe, because we’ve proven efficacy there and further reinforce that we’re a vertically integrated SaaS company. And that’s what’s providing this this visibility and insight that’s never been provided before.

Tony Zayas 34:11
would be curious to hear Katerina, if you see any other businesses out there that are maybe in other verticals, industries, they’re doing similar things to what you’re doing and taking, you know, these large data sets and understanding them to model human preferences and behavior instead, you know, you might be paying attention to others out there. But obviously AI brings in machine learning bring these fascinating new things to the table. And I’m curious to hear if they’re if there’s others out there that you’re aware of and what that stuff looks like.

Katerina Axelsson 34:49
Yeah, I’m in a way so there there are companies in the world of sensory science, but every every time I come across one I think it’s kind of similar, but it’s not quite like this. And usually the conclusion is it’s a partnership and not really competitive for us. And the reason is because there are companies out there that are looking at how do you formulate new products, for example, and they’re using neural nets to come up with new recipes. But we’re really what our technology is really doing is just telling you if people actually like that recipe, so so there’s a synergy there, right. So I don’t know if you’ve seen the AI weirdness blog or anything like that. But it shows computer generated recipes, for example. So sometimes it will come up with funny things like, get 30 pounds of cabbage in a mental leaf and stirred until it boils. And like, that’s funny. But what I’m saying is, TASTRY can actually tell you, is this a good recipe or not? Also, I haven’t any company that’s in the recommendation space, I feel like is different because they’re selling wine. We’re not in the business of selling wine. So I know it’s a long winded answer. But on the tech side, we’re trying to solve the same problem, but it’s not quite the same thing. It’s more of a partnership. And then on the on the business side, I have yet to see someone with a multidisciplinary business model like this. I wish I could I wish I could say yes, we have these competitors. And this is how we’re different. And this is how we’re faster, better and cheaper. But I think it’s just early in the market.

Tony Zayas 36:39
Yeah, and not even necessarily competitors. I’m saying like doing totally different things out there. But looking at similar things, understanding massive data sets and helping understand just human behavior and preferences.

Katerina Axelsson 36:57
So sort of, sort of so so there is a quality control software for beer, that I believe the company rebranded to analytical flavor systems. And what they do is they identify 24 characteristics of beer. Don’t quote me on this, but like how hoppy or sweet is it and they have expert tasters rate on a spider chart and how certain beers have those characteristics. And then they use AI to look at a new beer and predict if that beer is going to have those characteristics. I think quality control has a really important role to play in the current formulation. But again, I would say we’re in the business of telling you how many customers and where are they? And how much are they going to like that beer, not what flavors they’re going to perceive. That’s really the main difference in and the similarity is, you know, we’re working on product development and connecting products to consumers that will love them the most.

Tony Zayas 38:01
That’s cool.

Andy Halko 38:02
How do you actually engage with customers? Is it truly like a software as a service where they sign in and can self serve? Or I’m just kind of curious, versus, you know, how much do you have to get in with clients? And and, you know, go through scientific process and consult and help them through that?

Katerina Axelsson 38:21
Yeah, yeah, great question. We have. So when we’re selling the API or recommender, we have channel partners, so IT service providers and companies, like UST Global, who, who are already working with any major big box you can think of so they, they take the TASTRY API, and they go out and they help us implement it. And we talk to very major customers. So we don’t really we’re leveraging very large sales teams without with very little investment. So that’s been a benefit for us on the on the manufacturing side, we do have our own in house team. And the funny thing is, is we were iterating through the product, and they started coming to us. So um, we engage with directly with the winemaking team, and in some cases, the business and marketing team, depending on what what aspect of the software they’re interested in. So at the winemaking team, they approached us because they want to do AI assisted product formulation. And in America, better quality wine or more consistent wine or whatever it may be. And then the marketing or business team want to use this heat map data that we’ve generated to understand where’s their market where and where is their product going to succeed in the market, things like that.

Andy Halko 39:43
I’d love to know how you figured out that on ramp of working with other companies because there’s other founders out there that have a technology and they’re trying to figure out like okay, how do I get it out to the market, you know, was that the initial ideas that we would leverage these other companies as a vehicle, or did that come to life in some other way.

Katerina Axelsson 40:07
Um, for us, it made immediate sense when we came across the opportunity. And we just leaned into it. So we were discovered at the grocery shop, trade show buy this multibillion dollar company called UST Global and this, this guy comes up to me and says, we’re looking for innovative startups. And we’re really excited about what we’re doing, we think we can bring you a lot of customers in the retail space, let’s talk and then we talked and one thing led to another. And it just kept making sense every step of the way to lean into this partnership. And shortly after that, we partnered with another channel partner called eye control who provides inventory management services to retail, which opens us up to another client base and and they’re excited because they get to show their clients new, innovative product, and to show that they’re on the cutting edge of technology. And TASTRY is a part of that. It just, it just made sense. Every step. Yeah.

Andy Halko 41:12
No. And that’s great. I mean, I think that helps for other founders that are out there is that, you know, they call people lucky. But that’s a little bit of preparation and opportunity. Right. And so it sounds like that. That’s, that’s how it happened. And so I think that’s really interesting.

Katerina Axelsson 41:31
Yeah, I would, I would say, being open minded about opportunities that come across, you know, come your way that that served me well, is listening to people who have ideas and thinking through them. UST was one of those positive experiences.

Andy Halko 41:55
So how did COVID and everything that’s happened? You touched on it for a second, and people moving to e commerce? I’m both interested in how it impacted the business and your strategy, and then also your team and culture?

Katerina Axelsson 42:11
Yeah, thank you. Um, so right before COVID hit, our strategy was to continue continually grow our presence in retail. So we had, you know, produced some very impressive metrics on how we increase sales and make customers happier. And our store our rollouts, and deployments, where we provide pastry recommendations, whether it be website app, or on iPads, and stores, there were many mediums, all those engagements continue to expand. And we were predict projecting this exponential growth curve, as we were negotiating contracts with our clients. But when COVID hit, they started to really focus on you know, keeping their customers safe and keeping essential is on the shelf. So safe to say that exponential growth curve and retail did not happen. We renewed our existing contracts, which was good, and their conversations have come back with a vengeance since then. But we essentially have about a nine month law in 2020. With making progress in retail. I think every vendor dead. So what we decided to do around q1 of 2020, was take our 2021 strategy and just execute on it nine months ahead of schedule. And our 2021 strategy was to take all the data and insight and software we created and just repurpose it and sell it directly to wineries. Because it’s an opportune time for wine brands to be getting an E common this heat map is it’s predictive, like that’s the major selling point is we can see what’s going to happen before it happens. And when you’re getting into a new type of business model, you can say for the first time, that’s that’s a really useful tool. So we just we just quickly pivoted to, to or accelerated pivot isn’t the right word to vertically integrate. And it worked to our advantage. We we hardly did any sales or marketing. I think we put out a couple press releases and we on boarded 200 clients quite organically. So so going into 2021 We already had a presence on both sides of the supply chain. So I feel like I’m seeing 2021 is it’s growing. It’s accelerating faster than I had anticipated because of that.

Andy Halko 44:40
What about on that internal side with the team and we just had a founder on not that long ago that talked about you know, they lost a number of potential contracts as COVID came in. They had to lay off some people and you know, how did you how they were able to rally around that internally with the folks that they did keep. So internally, how did you manage through changes? And, you know, was there any, any change? And, and I guess, you know, culture and, and buy in from the team?

Katerina Axelsson 45:15
Um, I think they’re so so the buying in the team, I think it I think it’s strengthened because we saw, we were fortunate to have seen a very direct impact on our efforts to, to switch gears like that that came very quickly. For us wineries were very interested in the product off the bat, you know, and retail moves relatively slower, the sales cycle is much longer. So it was a surprise. So I think that definitely bolstered morale to see a direct impact on our immediate efforts, I think. We weren’t prepared for increasing traction on both sides of the supply chain, and the infrastructure of the company just didn’t have a chance to catch up to that. So right now, we’re working, you know, double overtime, to continue to accelerate product development so that we can deliver more services to customers who are demanding them. And that’s, that’s both an exciting and stressful time, because we have many major clients who are in some cases using version 1.0 of our software. So we’re excited to have that customer. But we also really don’t want to screw it up. Right. So stressful and exciting.

Tony Zayas 46:46
Katerina, I know you we talked a little bit about you know what you see three to five years out. But what do you have on the horizon here next six to 12 months? What what some of the exciting things you guys have going on?

Katerina Axelsson 46:58
Yeah, sure. We will be opening up a lab in Europe, and we will have tested over 3000 sques by q1 2022. And we are now expanding our efforts to also be in retail in Europe and our channel partners have a presence there. So we’re collaborating on that. So yeah, we’re moving outside of the US. That’s the major. That’s major next plan.

Andy Halko 47:31
That’s really cool. So, you know, what do you think over the years has been one of the greatest catalyst for growth? I think, again, for our audience of founders, is there anything that really helped drive growth, whether that was a certain hire or partnership, like you mentioned? You know, was there one thing that stood out as as a major catalyst for the growth of the company?

Katerina Axelsson 47:58
Hmm, that’s a great question. One thing to think of one thing, um, um,

Andy Halko 48:07
and if it’s not one thing, I think that that’s a good lesson, in some ways for some of the folks that that watch our show that it isn’t one thing. I’m always I am always interested if there is some turning point or some thing that helped click in place to make a huge difference. But if it’s not one thing, I think that that’s a great point, too.

Katerina Axelsson 48:33
Yeah, I would, I would say that, I mean, I feel like so fortunate that in q1 of 2020, we as a team made a very quick decision that we’re going to take this nine month wall and retail, and we’re going to turn it into a product development sprint, and we’re going to get our roadmap out, you know, execute on a roadmap ahead of time. That that was a major moment for us. Because since we’ve, we’ve done that we’ve proven out the business model, we’ve proven out that we can be a vertically integrated company, and that this is multiplying the value for everyone involved everyone that we’re working with. Huge I mean, because before that, we were talking about this idea of this data flywheel for years, but because we we made it happen, we can actually see it now and we can see it working. Yeah, that I guess that’s the best I can do. That was that was pretty major for us.

Andy Halko 49:42
that’s good. I mean, it’s the idea that you’re faced with an external influence. Yeah. Which happens the businesses all the time and then the ability to pivot on that and use it as a strength to kind of catapult you know the business forward in in a direction that was beneficial. So I think that’s a great idea for folks that are out there is that you’re gonna get hit with external factors you can’t control, and you do have to pivot, and if you can do it in the right way it can it can be, you know, not a negative as much as a positive.

Katerina Axelsson 50:17
Yeah. Yeah, I agree.

Andy Halko 50:20
It’s great. Um, so I, you know, one of the questions that I always ask folks, and this kind of plays in to what we just talked about, but if you were able to go back before the business started, let’s say back to your research days, and sit and have coffee with yourself, you know, what’s one piece of advice that you would give, you know, five or 10 years ago?

Katerina Axelsson 50:51
Oh, oh, my goodness. I would just say, be prepared to sacrifice. And stick with it. It’s so hard to know, sometimes how far along you are in the process. I don’t know if that kind of makes sense. Like when you’re 90% there, to achieving a particular goal or milestone, it doesn’t feel like that. Even when you’re really, really close. Sometimes you don’t know until you get there. And I’ve just seen over the years, so many brilliant startups give up just too soon, just just just too soon, just because you go in, you start a company and you you have this idea that you’re going to be at a certain point in a matter of years. Sometimes it doesn’t work out that way. Sometimes you have to iterate especially if you have a new technology that’s never been developed before and hasn’t been proven before. And you know, it’s a whitespace company, for example. So I would just say iterate, iterate until something works, but don’t give up. I think, yeah, a lot of people, I think, take the fail fast idea the wrong way. Like I’m not saying keep banging your head against the wall if something’s not working, but don’t give up too soon, either.

Andy Halko 52:25
Yeah, we’ve heard that phrase fail fast a lot. And I tend to like it. But But you’re right, is that, you know, it’s about iteration and not necessarily stopping and, and reinventing. Yeah, if you will. So that’s, that’s great. You mentioned the word sacrifice. And something that came to my mind when you said that was we always talk with people about work life balance. So for you as an entrepreneur, how have you found for yourself don’t just and work life balance to me is always weird because if you’re passionate about it, like it’s just part of your life, but maybe how do you handle the the mentality of, of, of, you know, building a business that is can be stressful? There’s ups and downs, you know, there, there is a lot of hours, how do you handle the mental side that keep yourself, you know, engaged and focused and driven?

Katerina Axelsson 53:23
Yeah. I’m still figuring that out. Um, it’s difficult. It’s hard to turn that off switch, we’re on off switch off, right. I realized, though, that if you don’t do that, you will be less productive. And you will not make the best decisions if you don’t have a clear head, if you haven’t had a chance to relax and rejuvenate. So I personally, I love to love to before COVID go to the sauna. I love to exercise I have to exercise every other day to say, you know, positive, you can say and I used to feel guilty about that. But I schedule it. I have a block off in my schedule this like rejuvenation time, and I just consider it an investment in my productivity. And that makes me feel a lot better about it.

Tony Zayas 54:26
That’s great. So, Katerina, before we wrap up here with a last question for our viewers, you know, where can they follow and see what both you and TASTRY are up to?

Katerina Axelsson 54:41
Yeah, um, you can while we were present on our website, not so much social media yet. So I would say if you would like to reach out, I get forwarded emails from [email protected] and I’ll likely reply the second best option is on LinkedIn, I’m fairly active on LinkedIn.

Tony Zayas 55:03
Very good.

Andy Halko 55:07
So with tech founders, one of the things I always want to hear about, you’re on the bleeding edge of new ideas and just, you know, tend to be very innovative. What do you think across all the technologies that are out there and that are emerging? What do you think’s going to be most impactful for us as a society over the next five to 10 years?

Katerina Axelsson 55:31
Oh, my goodness, that is a great question. Um, I, I’m really excited about what’s happening in the biotech slash bioinformatics space, I think, you know, people talk about the Law of Accelerating Returns, but you really, really see it there. I think there’s going to be huge impact on on medical and health. And I don’t know if you’ve recently seen any info out there like by David Sinclair, and his new study on rejuvenating the eye cells of mice like, and setting back the biological clock with a gene therapy. I mean, that would have been unheard of 10 years ago. So I hope I answered your question, but a lot of progress being made there.

Andy Halko 56:28
So you’re telling me we’re all gonna live forever?

Katerina Axelsson 56:31
I think there’s a no, that’s not what I think that’s clear. Thanks. But I think there’s a 50% chance I really do believe it.

Andy Halko 56:40
Pretty interesting stuff. Yeah. It’s, and then you add in artificial intelligence and all these other things, and it’s kind of interesting, more this a lot. Yeah. Pretty Cool. Well, thank you. That was awesome.

Katerina Axelsson 56:55
Thank you so much.

Tony Zayas 56:57
Yeah. This has been fantastic. Katerina, thank you for your time here today. Thanks for everyone who tuned in. We will see you again next time. And everybody take care. Thank you, Katerina.

Katerina Axelsson 57:08
Thank you have a good week. You too.