Spark Tank - Edition 5

Welcome to another Spark Tank. This is the 5th edition. We have 5 founders pitching 4 VCs today, probably a bit more than 5. I think it's for 7 founders today. So it's gonna be fun. The founders have less than, uh, 3 minutes to pitch their companies to the VCs. The VCs have less than 2 minutes to give feedback, ask questions, and at the end we decide of a winner roundtable style. So I'm just eager to welcome, uh, the contenders. So we're gonna start with the founders today, quick presentations about themselves and their VCs, and then we're gonna jump right into first contender Iliam. So Sudan, welcome to the show. Tell us a bit more about yourself.

I'm the founder of Nicki's, uh, I founded the business based on helping parents with anxiety with nappy change.

Nice. Norbert, quick intro about yourself.

Hi, Norbert Herman, I'm from Lively. We're the cure for queuing. I'm based in Australia and we're solving the problem of delivering better experiences within stadiums, venues, and events.

Nice. Axel, a bit more info about yourself?

Yeah, my name is Axel Kloth. I'm the founder and CEO of a semiconductor company, Abacus Semi.

By education, I'm a physicist, computer scientist, and I've been working with supercomputers and designing for them for about 40 years, a little bit Nice.

Evan, a bit more about yourself?

I'm Evan with, uh, Kumator AI. We are the nervous system for supply chains, helping manufacturers deal with stuff like the closure of the Straits of Hormuz.

Ilya, a bit more information about yourself?

Hey, I'm from New York City, Ilya Zeldin. I am with ToKnowMe, which is the personalized professional learning platform for teachers.

Eric, second time on the show.

Yeah, uh, hi, I'm Eric Burgess. I'm from Los Angeles, California. I'm a creative technologist, um, and I am the co-founder and CEO of Credtent, which is the, uh, it is the trust infrastructure for the AI economy.

Nice. Shubhantar, uh, tell us a bit more about yourself.

I'm Shubhantar. I'm with Devish Mitha. We actually work on industrial automation, uh, platform development for, uh, for steel and mining, mining sector.

There you go, husband and wife team. There you go. Let's go with the VCs now. Um, we're gonna take the founders off stage and start with my good friend and co-founder Stefan. Quick intro about yourself.

Hey, I'm Stefan Whitwell. I'm the CEO of a nationwide wealth management firm and multifamily office. Most wealth firms manage money. We help clients design freedom. I'm also a co-founder in the AI space, and I love backing ambitious people with real substance. And here to see who's got the clarity, the grit, and the leadership to turn a vision into a great company.

Sunny, quick intro about yourself. Co-founder and friend as well.

Hey, what's up everybody? My name is Sunny. I'm a co-founder of GTM Ventures, Raz Rocket. I also started India's first Bitcoin exchange 13 years ago called UnoCoin. Uh, we help the daring build legendary companies. Excited to be here.

Marianne, uh, second time the show. Quick intro about yourself.

Uh, I'm in Berkeley, California, and I work for Future Frontier Venture Fund and the Berkeley Gateway Accelerator. We work mainly in, uh, deep tech, robotics, AI, and material science, and, uh, industrial manufacturing, biomanufacturing in particular.

Perfect. So that was the VCs, the panel. Oh, a quick intro about myself, going to be VC side as well. Co-founder of Mactan, uh, we invest in YC companies, $10 mil YCBatch, YC application is soon, so quite excited about everything that's happening. Let's have our first contender on the stage. We have Ilya up first, so it's gonna add you to stage with Tonomi. And yeah, we can see your presentation, we can hear you loud and clear. The stage is—

I thank you guys, everyone. So welcome to Tonomi. We are the operating system for professional growth. There are 3.5 million teachers in the United States, and that's a $4 billion addressable market in K-12 alone. And their professional growth is still stitched together from disconnected tools because today the districts, the school districts, they cobble together 3 or 4 different tools for goal setting, observations, and professional learning, but none of them really talk to each other. And so the result is compliance theater not actual growth. And this inefficiency has been priced in for years. But the landscape is shifting because you have states that are mandating structured evaluation frameworks. Teacher burnout is making professional learning a retention issue, and districts are under pressure to adopt AI, but they need AI that drafts and organizes, not the AI that scores and judges. And so we sit at the center of all three. Uh, Tonomi replaces that patchwork with one connected system. Teachers set their goals, admins observe and coach against those goals, and the platform surfaces the right learning to close the gap. One cycle, one platform, one shared language. We've gone from a few hundred seats to over 9,000— 8,000 in 3 years. We are in 120 school districts, retaining 90% year over year. And a fast-growing pipeline heading into next year. The model is built for compounding. You have 90% net retention of revenue, 80% of our contracts are multi-year contracts, and a trajectory that sets us up towards that $5 million in annual recurring revenue based on the current use case. Uh, we have a strategic partnership with the Danielson Group, which in this case is the most widely adopted instructional framework used by 60% of our school districts of our market. And that means that we already speak the language that districts use every day. Uh, once on the platform, the stickiness value is enormous because all of the workflows live in one place in the Tonomi system. The team has been in edtech and educator growth for years together. We've built the product, we know the customer, and we've been selling it into this market since before it was trendy. And I think the cool thing is that this playbook for teachers is repeatable. We partner with the framework expert, scale adoption through network, and measure outcomes. We're proving it now with public and Catholic schools. Next will be workforce skills and nursing, and that's 3 to 4 ARR multiple from the same engine. Today we're raising, uh, half a million, and every dollar goes to the go-to-market because the product is built, the model is proven, and we know exactly what works. And so now we need to accelerate. Thank you very much for getting to know me. We can improve professional growth for adults together.

Nice, love the crisp presentation. Steph, any feedback or question for Ilya?

Yeah, I was curious, how do you measure success?

Uh, the, the most direct way is sort of observation or evaluation with teacher before and after. Right, what was for a particular set of skills. Uh, a more difficult but more— but a better way is to actually start measuring it through student outcomes. And a proper way is to combine the two together, because quite often you see one measure through the lens of the principal or a teacher themselves, and a quite different picture when you, when you superimpose it to the actual outcomes of students.

Gotcha.

And when do you anticipate being able to hit breakeven?

We are at breakeven now at $300K a year.

Fantastic.

So that's why, that's why I'm just looking for money to grow, to hire more, to just help me, uh, be that, uh, market-facing go-to-market.

Yeah, no, listen, I think you did a really great job at presenting, like, uh, Charles said, in a very crisp way. Um, I would maybe highlight the fact that, you know, you're already profitable, which is pretty impressive considering where you're at in the stage. Um, with 3 people though, I don't know, you know, that, that, uh, I mean, you're gonna need to ramp up. And so I think just giving a little bit more color in terms of the, the vision as a company. Um, are you looking to build this into kind of a broader company standalone long term? Is it going to be kind of an exit that you're going to sell to a strategic? I think it'd be really helpful to just kind of complete the story. But what are you raising, equity or SAFE, or what are you looking for?

Uh, we're raising equity. We're almost, almost there. But, um, yeah, just great.

Okay, so then maybe mention just how much more you need and what kind of valuation to kind of get the listeners here engaged. But I think a great job.

Thank you very much. I'd be happy to follow up with those questions that you asked because they're valid.

Yep.

Sunny, the architect of the deck.

That's true.

I, I must credit— wasn't sure if this is the right venue, but yeah, I must credit Sunny.

Yeah, yeah, yeah, go ahead, Sunny.

Uh, no, I was gonna ask you, actually, I know we've spoken before, but how big can this get, uh, when you consider just the core wedge alone?

Uh, we, we'll look— so the 3.5 million teachers were really the, the addressable market is, let's say, half of it. So we're looking to have 140,000 teachers under management, and we have about 10,000 right now. Uh, we're looking at $5 million in recurring revenue at roughly $35 per teacher per year. So, and that's just one example, one expert whose expertise scales wide. Uh, if you think about other experts who have frameworks, who have methodologies, ability to scale the quality and implementation fidelity is why we think we we have a real chance here.

And then, okay, sorry, Charles, just one last thing.

What's up? Go ahead, go ahead, because Marianne is gonna not join us on this one, she's gonna join on the next one, so you can take a bit more time.

What's the hardest thing to copy in your product or the data layer?

The, the, uh, the hardest thing to copy, it's a good question. I think, uh, A lot of the tools are built for specific needs, and it's hard to build a real universal, almost like a platform that, um, that is applicable to such a diverse and fragmented market. That from the very beginning, from your architectural beginning, I wanted to apply my technology background skills to create that type of a scalable environment because it comes last, right? First is really the the understanding of adult learning theory, the social psychology behind the actual process of helping or making learning accessible and trustworthy, then the actual user experience— how do you create a visual story that helps people take themselves to the next level— and only then the actual technology architecture. And that, I think, that approach is hard to scale. Or let's just say that I haven't We're still uniquely positioned to really truly represent the lens of a user as opposed to a bunch of solutions that are kind of important, but they're siloed and fragmented and therefore not useful if you're just an adult learner trying to improve.

Yeah, great job, Ilya. I loved your presentation.

Thank you.

I have a bunch of questions for you, Ilya. I always start with the basic, right? The branding. To gnome is like— I was like, two gnomes? Like, why the Lord of the Rings reference? Like, why talk about this when we're targeting teachers? So like, what was the logic behind that? What's been the market feedback of that name?

Sure, sure. Uh, well, first of all, they're copying me. I'll put it that way. But, uh, the root of the word is really gnosis. Gnosis is Latin and Greek for knowledge. Over time in English language, the G dropped, but that really is, you know, if you think, go back to Plato, it's like, know thyself is the beginning of wisdom. So I want to create an opportunity for adult learners to become better aware, to know themselves. So that was really the root of the word. The preceding that, and after almost the official name of the company is actually BrandMe, right? That was the original name, BrandMe. And then I quickly discovered that as much as, you know, people talk about personal branding and all that, nobody wants to be branded, right? It's almost like labeled as something. So that's why the name changed to Tonomi. And I think the vast majority of people will bravely ask the kind of questions that you ask, but then ultimately say, no, no, the name is great, it's sticky.

Yeah, I'd like inquire more in my clients, but me, you know, as a sales guy, as an investor as well, I want to avoid as much confusing as possible. I want like clarity from the get-go. I want to make my stuff digestible, right? So I'd sort of question, uh, that. Then obviously I have a bunch of GTM questions, uh, for you, right? So what's your sales cycle generally? How much lead do you bring pipeline? What's your close rate right now? And once you get the money, like, how would that change?

So my close rate is probably 34% of what comes into the pipeline. The vast number of leads comes from our partners, strategic partners whose methodology we help scale, the Danielson Group. And at this point, word of mouth. We also have a pretty active kind of email nurturing situation, but that's it's not super productive. The thing that really makes this industry, I think, scale is relationships. And so it takes a while for the new guy to establish them. And that's why if, you know, if I were to spend the money and hire somebody, it would be somebody with relationships in this space.

Okay, so you'd hire your head of sales, right? Like, is that what you're saying? Head of—

Yeah, I would hire somebody in the sales and go-to-market space.

Okay, I just like one last question. What's the price per head, price per teacher? Give us, uh, economics.

Sure, sure. It varies from $60 per person for, uh, like a smaller size organization, uh, down to $15 per person for a statewide Idaho that we have.

Per month?

Per year. Per year.

Oh my God. Okay, that's probably too low.

Um, probably too low. I agree.

Sorry, but I understand like why you would price it that low because that's, you know, you need to play in like the range of $15 to $30 in that space. I've done edgy tech a lot. It's hard. It's hard as fuck to do. Go to market for you guys and the sales cycle are super long. How do you deal with that mentally as a founder?

With degree of difficulty and alcohol, but It's working. I'm trying to open the space up for small and mid-sized school districts as opposed to enterprise organizations. They're different kind of market dynamics there. But the sales— and one of the reasons is that the sales cycle is much smaller. The kind of decision by consensus that's predominantly in education space, the consensus group is much smaller as well. So that's one of the tickets to grow faster while it takes time like those 18 to 24-month sales cycles to percolate it with really, really large districts.

Got it. Well, thanks for, uh, showing up. Yeah, we're gonna go to the second contender. Um, we have, uh, Sutan on the line here. So hi, Sutan. Um, can you hear you? We can see your presentation. The stage is perfect.

20 days, right? My son was in intensive care and he had an infection. And that nearly killed him, right? Um, me and my wife brought him back home and we looked at every Baby Wipes and every baby product that we're using, uh, and found out all the chemical toxic preservatives in there. So we decided, would we put this on our child? And we said no. That's how the business essentially started. Um, so this is the key products that we have. So the main product that we built was 100% natural baby wipes. They started from that. It's made of manuka honey and coconut oil, all natural ingredients. And the key aspect of our product is that we use less water. That's the key aspect. Everyone uses 98, 99% water. We use 89% water, and then we add all the natural ingredients in there— the manuka honey and the coconut oil. That took over 2 and a half years to develop. And we launched in 2019. And the key focus is one brand, one promise, zero compromise in the quality of the product that we produce. Some facts: we're 100% natural. We've actually sold 1.5 million products globally since 2019. We're currently, uh, forecast to 1.2 million in revenue. And profitability around $100K. Um, we've got 70% gross margin, we've got 5 new products launching, we've got 60% customer retention rate, and the market size is about $420 billion and growing. So we're in an amazing space we're in, um, and the product that we produce is that everyone loves it. And how can I say everyone loves it? This is the feedback that we get from our customers, right? The positive aspect is that we've built a product that everyone loves. We don't advertise. It's all the growth that we've had is purely come from, uh, word of mouth. And through that journey, we've experienced a lot. So we started in 2019, then we were hit with COVID Our product was stuck on the global logistics for 13 months. Like I said, our product only lasts 18 months from production, so we were burning money. We've gone into $1.5 million in debt. We recovered from that, and then we were hit with the global— with the tariff cost into the US market. Um, we survived that, we doubled down, and 2026 is our breakout year. We switched manufacturer, got our cost down from $1.80 to $0.62, so we can be more aggressive on our pricing and our growth. What I'm asking for— we're raising $1 million AUD, $8 million cap at a 20% discount safe note. We've already raised $600K and we've got $400K to go. Right, um, and the idea behind it, that within the next 24 months we want to reach 1 million homes, uh, get to about 30 million in 36 months, and trade sale or IPO within 3 to 4 years. And how am I going to spend the money? Majority of the money is going to go into purely stock and refurbishment, uh, and opening up the markets and operating costs and marketing. Like I said, As we built this business, due to the constraints of COVID and tariff costs, we haven't been advertising. It's all been word of mouth, and the growth that we've achieved is purely on customers' belief and the retention. And they stayed with us when we run out of stock, and that's the biggest testament for our product that has worked. Uh, and just cover short-term debt. Um, now what we would like to— what I— for me, this is not just a business, it's, it's a more of emotional connection. We've been through the difficult period, you can imagine, right? COVID, uh, tariff costs, and we survived all of it. And what we need now is basically cash just to accelerate our growth and a global expansion. We've got a product that works, parents who love the product, and we're at a point where the business is streamlined. We have 7 people operationally runs the business globally, and we're ready to expand.

That's it.

Thanks, sorry.

Yeah, I love it. I'm gonna start on, um, on this one, um, Maybe just a bit more practice on the pitch deck side of things, right? Me and Sunny can help you revamp the deck and just so that it flows a bit more, right? Like, just go through the slides quicker. I love the business. Why? I just had a newborn myself. We had a health scare on day 2, right? So I know how traumatic that shit can be. Like, I'm still traumatized to this day of, you know, like when you have this newborn and like all of a sudden their life, or you may feel that their life is in danger, that never sort goes away. So I love the story there, uh, maybe a bit more clarity like of what happened and how the wipes were responsible, right? Um, I'm a pretty obsessive buyer, like every single thing that I buy I research a lot, right? Including the baby wipes, right? And it's overwhelming as a parent, like all the fucking options that you have, what, what to choose, what not to choose. And obviously I went with the, the best wipes, right? We started with, um with Huggies and discovered quite quickly that this wasn't the best. I think we went with— is it called Water Wipes? Is that like your biggest competitor?

It's, uh, again, you broke off. Which one?

Water Wipes.

Uh, mate, Water Wipes is— yeah, it dries out the skin. We used Water Wipes, it dried up my skin. That's, that's why we ended up in hospital.

Huh. And that's how you got the infection? Dry this.

So what happened, the Water Wipes, uh, dried out the skin, which opened up his wounds, and then that wounds got infected. Which then spread to his kidney, and then that's why we went into hospital.

So the irony here is that WaterWipes, like, viewed as the, the best, right? If you do like research on, um, AI and all of that, they'll tell you. Um, so yeah, that, that got me concerned, right? I'd be a good client for you actually. I'm not sure if you ship to Mexico. Um, my first company was a nootropic, right, like 10 years ago, all natural. So, um, always been a biohacker and obsessed with the ingredients. I love that I saw coconut oil, manuka honey in there, which does wonders. Both are antibacterial, right? Plus coconut's good for the skin, honey as well. Maybe a bit more details as well on the why, right? And why this is game-changing. I believe, yeah, someone could do something with your brand. No bunch of VCs, consumer VCs. The economics were not appealing, but what I like about it is that it's a personal story, right? And me and Sunny, I always say it, we invest in founders, right? And in a founder, like, the iceberg that most people don't see is like the emotions. That's what moves you, right? Like, that's the full energy. That's what will propel you to keep on grinding even when the financials aren't there. So yeah, I don't have much questions about it, but I'll let Stéphane take the stage here. Steph has what, like, 4 kids, so been through there as well.

Yeah, I mean, what a great, uh, heart-led business. And, uh, credit you for having had the conviction to make it through that several tough chapters already. Um, I— a couple thoughts that come to my mind. So the first is you have a clear mission, and I think you've got a great size market, and I think if you price things correctly, you can have the margin to make it attractive. I mean, Charles would pay anything for the right kind of baby wipes, and so would anybody else. Um, and so I think it's important at this stage, at this critical stage that you're at, you've really got to think, got to have somebody on the team, or you have to make a little bit of a shift to really kind of up your business game. And that means you've got to back into and have really clear process on what net margin are you targeting, not just gross margin, but what net margin are you targeting. And how are you getting to that pricing? Um, you know, I think that, that too much at once is going to hurt you. So identifying just a couple markets, one, maybe two, to dominate first and then start accelerating to others— because to be dealing with tariff stuff at this stage of a company is crazy, right? Why not dominate one or two countries that are closest and then jump to the next one from a position of profit and strength? So I feel what you're missing is some really hardcore discipline from a business architecture standpoint. The good news is all that, that architecture is worthless if the underlying heart and core business and the potential of that business is not very attractive. You're sitting on a very, very attractive business, but you're going to need to add that discipline architecture now to that, I think, to really navigate the next couple of chapters. And it's going to be hard. It's going to be really hard saying no to other opportunities. Guys from Mexico, other countries say, hey, ship here, ship here. You've got to say no and have a super disciplined buildup. And then from the position of profit and strength, negotiate with stronger partners to help fund your distribution, to help figure out some of these supply chain issues, rather than trying to do that all at the size company you are right now. My, my two cents. I'd love to talk with you afterwards, uh, about collaboration helping you I think you've got a great, great business and a great heart.

Thanks. Yeah, Stefan just shipped me baby products, uh, from Austin, Texas to Oaxaca, Mexico, and they arrived 3 weeks later.

So I love to say that, um, we— Charles, we sell the products globally, but our key focus is Australia and the US market.

Yeah, APAC and MENA, right, would be Singapore for sure, Hong Kong, right? Um, could be interesting market. UAE, I feel, would be great. At UAE, man, that's like your number one market, like in, in luxury markets, right? Like people would pay for that. Uh, let's bring Sunny on stage and see what he thinks.

I got 3, 3 little ones as well. Uh, we have a 1-year-old, he sometimes makes a special appearance. Uh, but, uh, definitely hit home. Love, love what you're building. Love how it's heart-connected. I was going to ask you a question. So what does it take to get to a million homes?

Um, it's not that hard actually, because, um, from a B2C, direct-to-consumer base, right? And so at the moment we currently have 6,000 customers in Australia, right, um, that we service too, and kind of hitting that. So at the moment we're growing at about 2 to 3 new homes, uh, a month, because that's without advertising, purely word of mouth. Right, to accelerate that, uh, is like, well, estimate is about $40 per head to get into a new home from an advertising standpoint. That's a rough estimate for us to get there because the advantage of us is that if you look at the stats, 65% of parents do not change the wipes after using it from birth unless there is an issue. Our average life cycle of a customer for us is 3.5 years. Right? So we've got them. As soon as we get them, we got them, unless it's an issue, and we never had an issue, right? We have to convert that 30-odd percent to us. So does it convert? They're sold, right? Uh, and that's the advantage of us. Like I said, we don't advertise. We spend $500 a month on Amazon. That's it, right? We just started Facebook ad last month, uh, and Reddit a week ago.

Marianne, any question for, uh, so one, one thing first and foremost, I actually think you should be talking to David Yang in Hong Kong from Green Monday if you, if you don't already have your products there. Uh, very big inventor, owns a whole line of stores, very famous for selling vegan and clean products, and they get behind stuff like this and they've got a lot of, a lot of power in that, in that, in the Asian market sector. And then, uh, do these products have a shelf life or any kind of stability issue?

Yes, so we have an 18-month shelf life. The industry standard is 6 years.

Okay, great, that's helpful. And then, um, do you, uh, are you certified organic at all, or have you tried to get that?

No, we're 98% organic. So the way the wipes works is that if you have water, you— if your product has more than 90% water, your class is organic.

Okay. Are there any issues with honey and infants under 12 months that impact the use of this product?

Only at consumption.

Okay.

Okay.

And then I guess the other thing I was going to ask is I do a lot of this circular economy stuff where it's like leveraging waste streams. And so are there, are there, are there ways in terms of what you're doing where there's waste streams that could feed your product? And meaning like someone's hauling coconuts or doing Palm or whatever that is, you know, like this is very hot with investors, this kind of like circular economy leveraging waste streams, or is your product, if you're not able to sell certain parts of your product, are you able to move that into a circular or a waste stream? I think that's super interesting in terms of what you're doing. And then the final thing is like, I didn't hear, are you intending to, because right now, like if I get an Amazon order, there's an upcharge because I'm in the US and it's $6 gasoline where I am in California. So are Are you— is your intention to build some kind of— if this is your largest market, are you going to build some kind of manufacturing facility in the US? And then how do you grow the team? What are your— how are you thinking of growing the team?

Yeah, we're in the US at the moment. We want to set up manufacturing facilities. We tried manufacturing in the US. It's too expensive. And part of the agreement and the cosmetic industry, we have to add a chemical preservative if it's produced in the US. It's no other way. So that's part of the requirement if it leaves the shelf. So we will never manufacture in the US due to those restrictions. So we manufacture in China and South Korea.

Okay.

And the circular economics, we source our own manuka honey because the cosmetic industry clarifies manuka honey below the grade that we need it for. So we source our own manuka honey, which we ship to China for production, and we source our own preservative, which we also shipped to China. All the other ingredients is brought in by the manufacturer.

I found the wipes on Amazon México, so I'll give it an order for sure. It's about, what, $60 USD here, which is, I mean, not a big deal for me. It's still a premium category. What's your retail plan? Last but not least, like, how will you get into these luxury markets in Dubai?

I mentioned we won't go into retail. We didn't go into retail because we just don't have the margin before. We do have the margin now. We're looking to go into retail, Costco in Australia. We've just been selected to go into Costco and we've just been selected to go to Aldi as well.

Right.

But we haven't gone to retail in the US because like Stefan said, tight focus, strict structure, purely B2C, grow that market, get the brand awareness, and then retail. It's easy getting into retail, but if I can't get the foot traffic and sell through, I'm dead.

Right, so yeah, thanks for presenting. We're gonna go to our next presenter, Norbert. Um, so Norbert, you have the stage, my friend. Uh, we can hear you loud and excellent.

Lovely to be here today. All right, I'm Norbert Herman. I'm from Lively, uh, the cure for queuing. So we've all been there. At that big stadium event, the crowd roars, that game-changing goal has just been scored, but guess what? You're still stuck in the queue for food. And it's not just you, it's also the vendors. The venues lose millions in revenue and lack critical insight from event to event. It's a billion-dollar opportunity screaming for a fix. So Lively is on a mission to deliver the gold standard event experience for patrons, vendors, and venues alike. How? Well, a little bit tongue-in-cheek here, but I like to think of it as Uber Eats for stadium seats, except we're purpose-built for large-scale environments. As per standard, patrons simply scan, order, and pay to get— and get notified when it's time to collect, or in dedicated locations, get it delivered directly to them. We also do pre-ordering as well, and it's not just for food and beverage, it's for merchandise, alcohol as well. Lively at the back end is the low-cost, high-return solution that plugs directly into existing hospitality workflows, and, and it's what they've been missing. It's all fragmented today. So we're built on an AWS enterprise-grade backbone, and we're not only highly scalable, we're secure and instantly deliver revenue uplift. We've already rolled out at advance in Australia and the Middle East and Asia. And out of the box, we, we see at least a 1.4x uplift in sales, as well as increases in repeat orders, less error rates on fulfillment and wastage. And these numbers improve further as we continue to work with our clients. 76% of patrons say that in-seat ordering would not only improve their experience, but what we've seen is that it uplifts cart size, repeat orders, and it has the added benefit of evening out the ordering flow outside of peak periods as well. It's a huge market. The opportunity is massive on all fronts with a conservative, uh, sum of $750 million for our beachhead customers in stadiums, events, and festivals. It is definitely a demonstrated tapped, untapped opportunity out there. And the market goes well beyond that, as, uh, into other areas like hotels, which we're also working into. So Lively stands out from the crowd both in low-cost deployment, a flexible white-label offering catering to most verticals out of the box, but importantly, our focus has always been on solving the problem first. It's not just a tech solution, it's a problem-driven solution. We keep the people in the moment, not in the queue, and deliver results that benefit the entire ecosystem. Here's just a sample of the events we've partnered with to date. There's some pretty big names there, but we've also got some bigger ones coming up this year. We've proven it works in market, and now we're ready to scale. We're also trusted by key clients with, uh, already signed contracts for 2026 in international and Australia, and our goal is to expand into Europe, UK, and the US as well. Our ask might seem a little low, but it's actually our starting point. We're asking for anything from $150K USD strategic investment at a $5 million pre-money valuation up to $1 million. And the purpose for us is really to expand our team, to get boots on the ground in the markets we want to go to, particularly sales acceleration and venue acquisition. More pilot, uh, deployment and scaling, some more tweaks to the platform, of course, and building out our data lake to feed in better insights into the platform, which is the underlying moat for us as well. We've got the right team to bring this to scale with experience across product, hospitality, events, and tech, with also a couple of successful exits on the books as well. Thank you so much for your time. I hope it's something you all relate to as well. I look forward to your questions.

I'm bullish on this, um, for multiple reasons. Yep, since COVID like, uh, people want to gather more at venues, right? Then I think it's not a problem that a lot of founders want to solve. So that's your little moat. It's physical as well, right? I've experienced that problem myself after going to a bunch of, uh, musical concerts and sports sports events. Um, my partner Sunny will tell you, I'm like, oh, let's go to the UFC, let's go to— we want to bring my clients through events, right? So we could even partner up. Um, but yeah, quite bullish on it. Um, my question is like your GTM, right? It's always around GTM on my side because I've, I've done GTM also for that niche, and it's, it's pretty hard to get in. Tell me about your, your main go-to-market funnel.

Yeah, so Rory, who's my co-founder, he's a machine. He's opened up a lot of doors for us and he's been on the journey with me for the last couple of years. So through him, just as some examples, we've secured contracts with Asahi Beverages. We're deploying with them here in Australia starting next month across multiple stadiums here. And the goal there is to get embedded into Australian stadiums and scale up from there. We've already done a lot of work in the Middle East as well. We work with Oakview Group, who have a $40 billion size portfolio and want to integrate us into their core service offering, not just in the Middle East, but more importantly in UK, Europe, and US as well. So what we're finding with that sales cycle is when you get those believers like Oakview Group and Asahi, they become launchpads for us to get into stadiums, to open those doors, and they have the, I guess, the sway and muscle to be able to take us with them on that journey. And I think a big one of the probably following questions you could ask is a big part of that for us, you know, we build for advocacy across every part of the ecosystem. So retention of these clients is why they're taking us with them.

Um, what is the competition out there? Is there like other enterprises doing that? Um, yeah, tell me a bit more about that part.

Yeah, well, I mean, one of the ones we can get asked about is Amazon Just Walk Out. Um, uh, you know, and this is why I sort of talk about we're solving the problem. Like, the Amazon Just Walk Out was designed for, um, supermarkets, right? The ones in Seattle and around. It's not really fit for purpose for stadiums. It's a million plus to execute. It requires staff to manage. It actually takes away from the concession holders and the venues, and people still have to get up from their seat and basically go do their shopping and then walk out. So it kind of defeats the purpose. It's a million plus dollar execution that doesn't actually solve the problem. We want to keep people in the moment, in the game, and not with the people that they love as well. Um, and I mean, we also have other advantages, uh, apart from like Amazon. You know, there's, uh, we— I looked at all of these things when I was building the product, um, from the ground up, and I think our financial model is very competitive. We understand that there are thin margins in the hospitality layer in these places, and so we have designed it in a way where it's easy to deploy, but it's also works to the benefit of the venues and the vendors as well.

Thanks so much for the answer, Steph. Any question for Norbert?

Norbert, what a brilliant niche. I don't know anybody that would prefer waiting in line, right? I mean, bring it to me. You see that, you see that in movie theaters, right? At least here in Austin you've seen that and it was really popular. But in big sporting events where the lines are much bigger, What a brilliant way to make for a better experience for the, for the people that are there. Question for you about the, uh, the competition in the sense that, you know, usually there's vendors that pay money or work hard to be like the provider of beer or food or hot dogs or whatever kind of stuff's being sold at the event. Um, are you using their food or are you bringing in other sources that compete with them? How do you deal with that friction?

Yeah, by default we're designed to actually work with them, not against them, and I think that's one of our competitive advantages too. Um, yeah, that's great. Yeah, and we, we have the ability for, you know, uh, venues— like, to be honest with you, whilst this is an enterprise-grade system, we've also done festivals where we've integrated with food trucks. Um, so we can do like the bottom end of the scale all the way up to the top end of town and because it is an AWS-based platform with that backbone, we got the scale. But yeah, our goal is to not compete, to support, and we're happy to be as invisible as we need to be within that.

Yeah.

And what is your margin like?

Yeah, so we— our goal margin, or like what we, what we've built to over the last couple of years, is about 20% take. Um, so that's our, that's our net revenue. So, uh, and we're on track this year to hit breakeven, um, with off the back of this Asahi deployment, um, that will see us bring in, uh, more than enough to break even, probably a little bit of profitability. Um, but yeah, that— and that's, um, and it's a numbers game for us, right? The more stadiums we're in, the more events we're doing, it just adds up very quickly.

Sonny, any questions for Norbert? Sorry, Steph.

Yeah, I was just wondering about stickiness. Like, once you land a venue, um, how sticky are they after they go live?

Uh, they love us. Uh, you know, they see the benefit. You know, it's, uh, it— I think one of the things, you know, because you only got a 3-minute pitch there, you don't— you didn't get to call out, but what we see is actually a decrease in, uh, front of house and back of house pressure because they get to control the flow. Versus traditional ordering mechanisms where you've got queues of people line up and you go, I want chips, hot dog, you know, and there's like a 3-5 minute cycle just to place an order for every person. And then payment doesn't go through, then you've got that. By the time we do it, you scan the QR code, all they get is the, the thing prints out, it goes to the back of house, and by the time the, uh, we're either sending a delivery runner if it's to a specific location, or a click and collect, that order's just there ready to go. And it's normally a 5-minute turnaround, often from 5-10 minutes from order to collection.

Robert, have you ever thought about offering coupons to help manage that flow? So like just during certain 15-minute periods or certain low, normally slower periods, offer coupons to help reduce that just crushing normal kind of aggregation of when people order together?

Yeah, yeah, absolutely. And, you know, like, we've, um, I guess, you know, the system is built to be able to get— whilst we do a progressive web app, so there's no software to download, Scan launches the app. Um, PWAs have come a long way over the last couple of years, uh, and, uh, there's a lot more you can do with them, particularly, um, in the, uh, Android ecosystem. So be able to do push notifications things like that. We've got a lot more control today than we did when I first launched. So we have the ability with this and with the data play, as we get smarter with that, to be able to say, oh, they're not selling enough hot dogs, let's do a push at this period. And they can initiate that and go, hey, 2-for-1 hot dog offer, um, you know. And so there's, there's a lot of that. But I mean, also the other side of it too, I think for us we're trying to We want to get in front of this. So pre-ordering, like you said with the movie cinemas, we want to be able to get there so there's a percentage of people that before they even get to the venue, if you're getting this on your ticket with your QR code, we can integrate with ticketing partners to enable people to pre-order for things like drinks and merch. So that helps the vendors to better plan their SKUs and volumes even weeks in advance. And we've seen this like when we did, say, the UFC in Macau. Our biggest headache was that we sold everything out before the first foot hit the first face. So it was very successful, but we don't want to be constrained by the SKUs either. So we tend to see like huge demand there.

Marianne, any questions?

Questions and comments. One, yes/no on alcohol.

Yes, we have a— I guess we call, uh, we call it an RSA compliant system. Okay, so it basically prevents people from over-ordering, it prevents people from, uh, repeat ordering when we can—

I want to go fast, so let me just— good. Uh, have you thought of surge pricing like an Uber? I think that's interesting. And then, um, uh, I think you're losing out— this is just comment. I'm the only female on the panel today. That's not always the case. You don't have a female in the mix of your organization. And one of the things that— like, I go to a lot of tech AI conferences, and the bathrooms are lovely, and there's no line for the ladies. However, the men's bathroom had the long lines. And at venues and events— like, I was at Rugby World in Las Vegas a couple of years— last year. And the line— so the trade-off that you could have a slide if you had a female partner in the mix and you had a slide that that person was speaking to saying like, look, the trade-off is waiting in line for the loo or getting a snack or a beverage or whatever. And I think that that is, that's a very, that's a reality of many of these venues. And then final question is, so I'm at a big open-air concert, how are you finding me?

Yeah, so we actually did say the F1 Concert Series, we have signage not just at the Heineken bars that we were working at, but around the venue that made people aware that scan the QR code, place an order, you can get food, beverage, merch, all of that sort of stuff. So we work with the clients on signposting and pathfinding.

And you OEM in with the vendors? So you're just like Lively inside or enabled by Lively versus being the brand per se?

Yeah, we're not fussy. We can do either or.

Yeah, gotcha.

There you go. Well, thank you, Norbert. Let's go with our next contender. We have Evan that's gonna take the stage. So, Evan, you have the stage here. We can see your screen. Hi, go ahead with the presentation.

Hi everybody. Yeah, so I'm Evan, co-founder and CEO of Keymoter AI. We are the nervous system for supply chains. What that means is we sense disruptions like the tariffs that we were talking about with the previous presenter, or disruptions like the closure of the Straits of Hormuz, and then we help planners deal with that so there's no delay to their production environment. Specifically, we help manufacturing planners make sure that they've got enough inventory so that production doesn't shut down, but also make sure that they're not holding so much safety stock that they're tying up all the their capital. And this is a global problem. You know, there's about $1.7 trillion that's lost annually to this inventory distortion problem—overstock and safety stock and wastage that comes from that. Then there's another $1.6 trillion in missed revenue every year from these manufacturing supply chain disruptions. And it's a huge waste of time. Planners, you know, spend about 50% of their time just scrubbing data instead of actually making decisions. So for our users, we've been able to save, you know, hundreds of thousands of dollars. For example, Sita here at Wilton Weavers— they're the lightweight fireproof aircraft— or sorry, carpet supplier to the aircraft industry— we were able to save them $325,000 a year, and that's a 6x ROI on our $54K subscription and 50% reduction in planning time. So the way this works is we have a patented causal engine that senses turbulence, measures whether it's going to be a ripple or a tsunami, traces that effect throughout the supply chain, and then runs simulations to verify the best action. That's based on 27 years of causal AI research, and it is patented. This is based on my co-founder's research. So he developed this while he was working at the R&D lab at Panasonic, and he did his PhD on this at EPFL. And I previously have been selling AI B2B systems like this for the past 15 years. I brought 6 high-growth startups to the Japanese market as the launch leader and developed a playbook for selling this type of technology. That playbook leverages a mix of direct sales— so the GTM is very focused for this year on the US across SF, Austin, Chicago— and channel partners for regions like Japan, APAC, and the EU especially systems integrators that currently own the tech stack. So we're projecting high growth over the next 5 years. We're already generating revenue across several household names like you can see here. Paid pilots are converting to ARR. We're targeting, uh, $100,000 ARR by the end of next month. And the pricing looks like this. Essentially, we sell an annual subscription, and that gives our users a volume of decision credits to draw down from. And for the largest enterprises, we take a share of the cost savings or revenue created. We augment rather than replace existing systems. So we're actually selling on the Microsoft network, augmenting ERP systems like Microsoft Dynamics. And we've just started the process to get registered with SAP. We plan to do the same for systems like Blue Yonder, etc. This category is called decision intelligence, and there's only one other platform that does this for supply chain. Called Aera. Uh, they've got about 15 years head start, but they're also slowed by being a legacy tech platform. So as a native agentic AI, we're much faster and leaner than they are. So we just closed our pre-seed round at $500K. Several VCs, early-stage VCs, came in on that round. We're opening seed rounds on May 26th when we do the Alchemist Global Demo Day in the Bay Area, and we're targeting a $3M raise which we'll use predominantly for GTM. So at least half of that is going to go to rolling out the US-based GTM over this year. The rest is to keep scaling up the product and general operations and so on. So to learn more, you know, please contact me, and I look forward to your Q&A.

Thanks very much. I mean, that was great. I'm bullish, right? I'll start with that. Um, maybe I would start with your background as well. Like, you're wild guy in Japan with like traveling on boats. I've seen like you're revamping an old car as well, your sales guy as well. That would have been fun to see. I'm bullish. The nervous system of— I'm not sure to use it as an analogy, you know, like I'm big on words and like nerves, you know, and nervous. But I can see like that even like skeleton is not better, but like yeah, I get what you're doing, but just maybe like phrase that, I would say. And why not start with tariffs or Iran, like, in the start of your presentation, like a real use case, right?

Yeah, can you answer that? Sure. You know, first, to my background, yeah, to your point, I think it's interesting because I kind of grew up in this industry. I started out in commercial fishing, and when I first came to Japan, it was for a supply chain use case. I was trying to help import Canadian seafood and better promote and sell it here in Japan. So a lot of time on the water, lots of time spent sailing and that kind of thing. Uh, I started a marketing agency to help promote Canadian seafood, and that turned into one of the largest digital agencies in Japan. And that's how I got a foothold into using this type of tech for AI recommendation algorithms, big data datasets and that kind of thing. So 23 years in Japan, it's been a very interesting ride. To your point about, you know, using the urgency of current conflicts and tariffs and that kind of thing, I think you're right. You know, a lot of people have a hard time understanding how urgent the problem is. We get a lot of questions about, well, why not use existing ERP or SCM solutions? So we're revamping that first slide to actually focus more more on the, you know, in-the-moment pain that a planter has. To the nervous system analogy, it's the ability to sense and respond in real time, just like the body's, you know, nervous system responds to threats. Part of the physiological reference actually refers back to my co-founder's PhD, where he was building basically digital brains based on the way our organic systems actually work. You can see on blog, for example, we talk about how we need to ground AI in the real world. We need physical constraints. We need to understand, you know, the size of packages that were recommended going on to trucks. We need to understand how much space is in the warehouse and things like this. So it's that combination of the physical sensing environment and understanding all these physical constraints together with this ability to have, you the generative AI, the world model combination, and the human in the loop, which is critical to our system. Yeah, so hope that makes sense.

Stefan, questions, comments?

Yeah, can you tell me, talk to us a little bit about your sales cycle or what your go-to-market strategy is when you, you know, when you first knock on the door to how you— to when you first get paid?

Yeah, at this stage it's all warm referrals. Um, so one of our advisors is former VP of Motorola, IBM supply chain, knows everybody in the industry. He's opening, you know, a huge number of doors for us. We're getting great results from our case studies, and so we're leveraging those to get introductions to more people in the industry. It's a pretty close-knit community, especially in the States. So for the first year, it's real kind of old school, in person, going to trade shows. Every customer I go, I do a tour, they're warehouse. I'd see the systems they're using right now. That's why the GTM is boots-on-the-ground heavy, and we need to scale up that physical presence.

How do you do proof of concept for a new client?

So we run a pilot that is on average 3 months long. One of our advantages is that we can implement very, very quickly using flat file data, right? So we're getting an export out of the ERP, we just get the bill of materials, which is usually a spreadsheet, and then we can build out our entire model, connect external sources, and build what we call a causal decision diagram in a few weeks. Uh, so in month 1, we're getting the data. In month 2 of that pilot, we're getting human-in-the-loop feedback. Uh, what we do is a continuous learning rather than a pre-training system. By month 3, we're showing ROI, and then that wins us the stakeholder alignment to then go through a vendor procure process and work with the IT department to, you know, build write-back APIs and so on.

Sunny, any thoughts, questions?

Yeah, I was gonna ask you, so, so two questions. One is how many ideal customers exist globally, uh, and then the second was around is the real product better forecasting or, or just like faster decisions?

Well, the core our product is a decision intelligence platform, right? So it's really focused around the decision as an object. It gets tribal knowledge out of people's heads into a record of decisions that we can then learn from and improve on. It's like Peter Drucker said, you know, what's— what gets measured gets managed. The problem with current systems is that it record— they record a lot of data, right? You can see dashboards and pretty forecasts and everything, but it doesn't tell what decisions actually led to an improvement. So that's a core part of the platform. I think you had a two-part question there, Sonny. Remind me, what was the first part?

How many ideal customers are there out there?

Right, right. So our ICP is manufacturers that have $50 million in revenue and up, up into the billions. It's a good fit for large enterprises that have complex supply chains. You know, our ideal customer customer is like SpaceX. You know, I'd love to work with the whole Elon Musk group. All of this— we're basically built for, you know, builders who are building like the Jetsons' vision of the future, physical, tangible improvements on, on our world. But globally, it's a huge market. Uh, you know, we talked about how it's a $3.3 trillion market. It's basically every allied democracy that's building tangible items.

Incredible.

I'll give the stage to Mary Ann. Any thoughts, questions?

And please, this is not like just— I thought the presentation, like, a little more energy would be a good thing. And I do, I agree with a few people, like, lead— I think leading with the problem would help you more. And also, um, I— maybe versus nervous, like neurons firing. I think that's something a lot of people can connect to, and that's an action. And I think that would, that would help you even in terms of AI. There's a, there's a link there. I'm curious, um, you didn't say anything about like catastrophic events and risk management firms, and it seems like that would be a market for you, but I'm not hearing you like the— that's not, that's not a market that you're thinking about. And also, I 100% agree. I think, uh, I'm in the Bay Area and you're— we're experiencing like this resurgence of tech companies hiring people because people don't wanna deal with AI agents and the lack of trust. It's just not there. And so you're seeing a lot more headcount hiring with straight up salespeople and engaging with customers and, and customers saying, I wanna I want an onsite visit. I want to see somebody. So I think you're spot on with that. Doesn't surprise me. And I'll stop talking now.

I mean, I'm sorry, I couldn't catch one of your points there. Were you talking about like that disaster response agencies?

Yeah. Yeah. Risk management firms, catastrophic risks like those big insurers that, that are always trying to de-risk something that you're like, hey, man, we've got these tools that actually are very predictive. You know, and I'm sure anybody on here is going to say like, what's going to prevent some smarty-pants kid in one of these organizations from just like building an agent and saying like, I'm going to scrape our internal intelligence and do something like this. So I'm sure you have a response to that, but it's something I think about.

Yeah, I think you have a good point there. It's also been pointed out that this type of intelligence is a great tool for investment bankers, private equity, and so on. We're focusing right now on empowering the manufacturers, and hopefully they can tell that story to their investors and leverage it. But yeah, I think that's a good point. For disaster response, we've seen a few solutions in the space that directly target that. Applying this type of technology to manufacturing is more of a blue ocean, and that's one of the reasons why we're focused here, but it's also our background, you know, it's the right, uh, founder market fit.

Thank you.

Well, let's go to the next founder. Uh, we have two more in the line. Uh, thank you, Evan. The next one in line, we have, uh, the husband and wife team, Sharanta. Uh, do you have this stage? Do you guys, uh, can you guys share your screen? Um, let's go with Axel till you guys do it. Axel, uh, the stage is yours, you can present. Perfect.

Yeah, so my name is Axel Kloth and I'm the founder and CEO of a company, a fabless semiconductor company named Abacus Semiconductor Corporation.

We build processors, accelerators, and smart memory to enable AI to be run not only in the backend, but also in what is called the smart carrier edge and where companies are relocating their IT back from the cloud into on-premises data centers. And as you can see, there is a big deal about these data centers because today the prevailing architecture with CPUs and GPGPUs, they're, are incredibly expensive and they consume enormous amounts of power, which is why Google now orders small nuclear reactors for its data centers. And where it's now said that about 40% of all data centers being built cannot be powered up when they are built, because simply despite the fact that power purchase agreements may be in place, there is no power to be distributed to those data centers. And so we got to do something about that because again the problems start to compound. It's data security, sovereignty, high TCO, low utilization, and just in general a lack of ability to build out those data centers at the rate needed for AI.

And, uh, of course that's not a good thing. As you can imagine, the TAM is enormous. It's in fact vastly bigger been even stated here, those numbers have already been deprecated.

So it's not $1.8 trillion, it's going to be much more than that. And we're at a stage at which something has to be done that improves the effectiveness and the efficiency of those AI data centers. And we're a company based in San Jose, California, global reach. The moat that we have is that our patent pending foundational architecture cannot be breached.

Somebody who violates that will have to pay, and as a result, we have a very defensible position with regards to our IP to make those next-generation server CPUs and accelerators to improve data centers. We can also build a data center in a box effectively and make sure that this can be used as a mini data center at the edge, carrier edge, smart edge, and therefore improving access to AI even for people who don't have AI coprocessors in their phones or on-prem. What we are addressing is really the rehomed IT for AI. Gartner assumes 47% of all corporations contemplate that. We're targeting agentic AI, And because of all of those are greenfield applications, there is no compatibility needed with x86 and CUDA. All of those data centers, unlike the ones in the backend, are very restricted in availability of electrical power. Most are in the range of 1 megawatt to 5 megawatt. I've seen one in the 10 and one in the 15 megawatt range and limits on cooling. And AMD and NVIDIA can't play there because of the technology. Technology. We seamlessly tile processors, accelerator, smart memory to give us the advantage that our patent protects and to be able to scale out near linearly. We have simulation data to prove that. You can imagine that if something challenging as this is being built, you need a team who's been there, done that. And I'm a serial entrepreneur, turned around HotRail from a $36 million loss to a $194 million acquisition in 10 months. After that, founded SSR Labs, Perimix, Axiado. Uh, my co-conspirator on the board of directors is the person who consolidated all of the Japanese microcontroller companies into Renesas America. Ever since that, they've been profitable. And then he founded Faster Systems and got it to a point where he could sell it. Uh, My CFO is a person with 30 years of experience. Our VP software engineering also 30 years software and firmware experience. Andrew Hsu has 20 years of experience. In fact, without him, we'd still not have touchscreens on cell phones because he was the one at Synaptics to actively push for them and got all the customers on board that now all make and build cell phones with touchscreens. So he's very good at conveying messages so that people understand what it is, where the advantage lies. And William Ang is our partner in Taiwan. He's the one who connects us to the server makers because we're just making the chips, or we have the chips made. Server makers will make the servers and the pods so that they can be sold to the users. And with regards to customer benefits, I can go through this, but our effective benefit is that we either maintain performance at half the power or at the same power we give you more than double the performance. You can go through all of those numbers, but anyway, for compatibility, we have the frameworks that we fully support, the same ones as everything else. Our security, the cybersecurity benefits are approved by the DOD. And if you look at what the market entry is, then we see with all of the threats coming right now and with all of the disadvantage of the high power consumption of GPUs, we can play in markets where AMD and NVIDIA can't. So agentic AI server farms, mass storage appliances, and network edge proxying. Uh, we have pretty good traction.

We have, we have traction so far—

WWT, Foxconn, and so on. So you see that we haven't been sitting on our butts. We have access to plenty of patents. I wrote about 40 of them. We're looking for $30 million. Most of them will go into product development, some of that into team building and the market expansion to evangelize our IP.

And in more detail, this is what it is.

And I'll go into the appendix a little bit more. So if you look at a typical Gartner quadrant thing, you can see if you look at the level of parallelism and then the effectiveness of the scale-out and the efficiency, we're very clearly at the right-hand top versus NVIDIA's a bit further to the right there. More parallel, but they have no scale of efficiency.

And if you look at CPUs, they're all in the left bottom corner because they don't have enough cores to deal with the problem that is impeded upon us by the use of AI, particularly agentic AI.

And that's it. Thanks, thanks, Axel. Second time on the show. Out of the 40 patents, how many are granted granted, vs. how many are pending?

So out of my 40 patents, one is at this point in time still waiting for to be granted.

All the others have been granted and they are usable and accessible to us. And the most important ones are of course the one that is currently in the queue, plus the 8 ones that I wrote for, for my prior company. But upon exit I managed to reach a deal in which we can use those patents for free, free of royalty, free of any kind of license fee.

I want to do something special here. Evan, um, can you hear us? Or, um, tell us a bit more about the supply chain and the potential disruption, uh, with everything that's happening in the world right now with chips. Uh, I guess both of you can speak about that. Evan, any thoughts on that?

Yeah, there's a huge amount of disruption. I mean, this closure of the Straits of Hormuz means that helium can't get out of Qatar, and helium is critical to chip production. So there's going to be, you know, huge knock-on effects from that.

Correct.

And the other big stumbling stone right now is that a lot of data centers are being built in the UAE, and in fact in most of the Middle East, and those data centers will not come online anytime soon because of the restrictions to get materials there, to get processors there, to get servers there, and even simple things as generators and cooling, right? So since they're not getting there, we're now basically constrained to anything we can build in most of Western Europe and in the US. And that means that compute is going to be even more restrained, and there's even less compute available for more and more demand, which grows not linearly but exponentially. So the exponential growth of demand and the cutting off of data centers that had been either in the pipeline or have been built in the Middle East, uh, that poses a problem that now makes us necessary to rethink the way we do compute.

So is that a pro or a con for you? How would— yeah, because I, I would view that as a retardant to to get, uh, yeah, the actual product to market and potentially even getting the capital. Is that a concern, or do you have a solution for that?

No, that's a pro, because ultimately, if you think about it, if we scale better than NVIDIA, we only need half to maybe a third of the components that NVIDIA needs, right? So if you can get the same amount of compute with half of the components, then you're not taxing the supply chain the same way that would as if you went with NVIDIA, right? And one of the, the big disadvantages— NVIDIA's processors and GPUs themselves, they're not all NVIDIA IP. So if you look it up, for example, there's a lot of stuff under that's called HBM, high bandwidth memory, that's made by SK Hynix, and that has to go through the same supply chain. So now you have multiple components going into one chip, and they're all dependent on each other. So if, if NVIDIA can't get the HBMs, then NVIDIA cannot make and sell those GPUs. And we don't have that problem because we don't rely on these GPUs and the HBMs the same way that NVIDIA does.

In fact, we don't need a GPU in our system at all.

We, we can build the accelerators that are much more tailored to the math problems that AI needs to solve than GPGPUs.

Evan, any comments on that? Or—

yeah, people tend to underestimate just how hard it is to build physical products in, in the, the real world. You know, we, we're kind of used to the concept of the world is flat and globalization, but the reality is, you know, just getting rare earths, uh, getting things that we track a lot copper and neodymium and raw components like that. It's very, very difficult to actually get those materials in order to build things. And so anytime you can reduce the complexity of what you're building, any technological innovations that allow you to be more efficient with your actual physical materials obviously are a huge competitive advantage. So yeah, it sounds like you're on an exciting track.

Stefan, any question?

Yeah, I was going to say, who would your ideal partner be, Axel, and how much money do you need, uh, to get to, you know, to where you can really effectively meet?

Yes, ideal partner would be a server maker that has access to our customers' customers, right? Because ultimately, uh, they're on, on the planet, there are 12 server makers left. They're all in Taiwan. And if you buy something that says Dell on top of it, well, it's— guess what? It's not made by Dell, right? And the same is true for HPE, and the same is true for Lenovo. They're all made by others. It's for the most part Quanta and Wistron whiteboxing stuff for all of the guys, right? So Quanta and Wistron in Taiwan, they make— I would say they themselves, these two alone, make like 50% of all servers, right? And then if you add to it the others like ASUS, and then literally all of the servers on the planet are in Taiwan. Now there's an additional complication. If you put a GPGPU in, you have to rely on very high layer count PCBs, which are made in two locations. There's one fab in Germany making up to 60 layers, and there's a whole bunch of fabs in China that make them 30 layers, right? In the US, you won't find anything above 12 layers. So that means that if your server has GPGPUs as a, as a requirement that are built in and you can't get it down to below 12 layers, guess what? You now are waiting for the processors and the memory and the GPUs to come in, plus the PCB that takes all of that to be coming in from China, right? With our case, the PCBs we can use because we all use point-to-point connections, and we don't have the need to go for very deep stacking of layers, we can easily deal with the 12-layer PCB, meaning we can completely circumvent the issue of a bottleneck from PCB manufacturing at 30 layers in China. We don't need that, right? And so that was a big part of the, the consideration when we initially architected all of this, because it makes a whole lot more sense to have the complexity in the chip chip, actually not even complexity, but the features and functions you need to connect them in the chip versus on the PCB, right? And so that's why we have lots and lots of benefits, not only performance and power but also ease of use, ease of manufacturing, everything, right? And so we believe this is in our benefit, and we believe that we have a much better outlook on the future of AI because the future of AI is already trending away from the massive scale of training only and then inference somewhere else. It's all going to agentic AI, which means, yes, you still have one incident of training and creating your, your foundational model, but then everything else gets updated through an agentic AI instance that now needs both the flow control and the math required for the update of the model. And we incorporate that because we can tile the processors together, which NVIDIA—

there you go. Sunny, any questions for Axel?

Yeah, Axel, you, you've spent, what, 4 decades designing supercomputers. What do you know about this, this specific problem that, that nobody else does?

Well, in a lot of ways it is experience, because all engineering problems come back at you, and the issue is only the pendulum swings, right? So supercomputers first started out as really unique. Every single one was unique, right? And they were all hand-built and they were extremely expensive. And if you look at it, if you've got a phone today and if it's modern, reasonably modern, right, your $1,000 phone today has more compute power than a supercomputer out of the year 1989. And that thing back then was a Cray Y-MP cost about $10 million, needed 3 megawatts of power, right? So the, the issue is the, the pendulum always swings in between 2 or 3 extremes that you have with regards to what your design focus needs to be. And I've seen it, and I know the problem is that most people think you can feed in power, that's the, you know, that's the biggest problem problem of getting in power to the chip. No, the biggest problem is to get the heat out. And for the heat to get out, you have to have a more efficient compute. And for efficient compute to enable large-scale problems to be solved, you need not only the performance of one core, you need the ability to have many, many cores talk to each other at the same time in a way that is unrestricted, where there's no collision and where contention is managed, right? And that's what I did because I've been doing this and following this industry for so long and developing for it and using it, right? I use supercomputers and I still rip my hair out.

We should start the first slide with that. Then I'm saying, hey, uh, Moore's Law stopped scaling. Do you agree with that factor? And how could we not, like, start the deck also with this?

Uh, well, so I mean, to some degree, Moore's Law of scaling still, well, to some degree applies, right? It used to be that every, every 18 to 21 months we were able to double the number of transistors per chip, right? And that we achieved that by making the chips bigger, and we achieved it by getting better density. But that's long gone. We don't double it anymore. We maybe get to a 10-15% increase in the density of transistors per chip. And we have now reached radical size, meaning we have reached the size at which the machine makes the chips as a single entity. Right? And that's 28 by 32 millimeters. Can't get bigger because that's really, you know, the, the photography machine size limit. It doesn't get bigger, right? So since we're at that limit, it basically says that the only thing we can do is, if we can't squeeze more transistors per square millimeter and we have limitations of theoretical, then the only way to scale out is make sure that those dies talk to each other effectively and efficiently and with as little energy as possible, because energy wasted means again heat production, right? And if we're at a stage where heat is anyway the issue, we can't get rid of the heat, then efficiency becomes the paramount.

And I feel that's what we need to explain the first slide clearly. Marianne, any question, uh, for Axel?

Uh, Move the team slide to second and clean that up. It's messy. Say the problem you're solving first. And also, the best gift anybody ever gave me was, "And the short answer is..." And I think that would serve you well. You've got the answer, but I think you need to tighten it up and say, "And the short answer is..." And then I don't know any direct person you should speak to, but you're in the heart of Silicon Valley. I don't know why you're not shopping this there. I do think for Spark Ventures has an appetite for, you know, what you're selling here. So more feedback than questions. I don't have any particular questions, but I, I'm a little curious why, based on where you're located, that you're not shopping this elsewhere. Um, the short answer is—

and the short answer is we've tried a good portion of Sand Hill Road.

So, okay.

Yeah, thank you, Axel. Great, um, great company again. So let's go with our last contenders. Um, Shubantra, yeah, you guys have the stage. Maybe move closer to one another so we can see both of you. And, um, hi. Yeah, we have the— your deck here.

The stage is yours. Okay, so we are working on— I'm Debasmita. So, uh, we are addressing the problem of an AI platform. It's IndusOne Industry 4.0 solving steel plants' downtime losses is very huge. We are working on that to resolve it. It's $1 million per hour. So that's the problem that we are solving. So here the steel plants are honestly, they are blind It's due to the legacy issue. The plant's infrastructures are really old. They don't have the real-time visibility every time. And also the issues are the large-scale manufacturing facilities to change also takes a lot of money. But the problem that they are facing, they are very, very conservative. So the problem that they are facing due to unplanned downtime, which is a very huge number, $1.4 trillion. So they think it is not solvable, but these are all— 80% of them are predictable solution and it is easy to resolve. So we are addressing into that problem. So here we are trying to— it's a three amount of forces which are coming together. One is the highly computation intensive AI/ML work at the edge, the Second part is the connectivity and the IoT supplying this and resolving the supply chain issue of downtime for the steel factory, which was long before. Now it's, it's not possible for them to really wait. So the— here, uh, the total number of conveyor belts is 1.8 million and the market size is very, very huge. So we are— the each conveyor belt is 100K per year. So the market across globally is 240K. So which is very huge. We are addressing just 5%, which is also 13,000. This is from the Siemens report that 2034 the CAGR is 16.3%. It's double digit and the market is $800 billion, which is presently at $239 billion. So we are— so this is the problem we are solving: capturing the data, Industry 4.0, through cameras and multiple sensors, analyzing them through computation-intensive hardware at the edge edge and getting the solutions or the results of the analytics and very quickly in 0.3 seconds, sometimes in milliseconds depending upon the, the how the speed of the conveyors that and the usage. One minute left. Okay, so, uh, we are solving this. This is, uh, POC's problem we have resolved to in order to get 27.3 $30 million profit benefit across steel market. This is for steel different shops, and this is true to two means. We are giving the hardware device as a solution, with that we are making 30% profit, and the software that goes, uh, for the AI/ML platform and the dashboard, this software licenses goes on for 5 years. So that's why, that's where I said, 100K, there's a 50% profit coming out of that. And so coming to the competitors, so we are, we have a huge model which has been already built with their failure models, and that is where we have the platform ready, whereas the competitors have datas which are coming from PLCs and very old systems which will not resolve the AI/ML model actual data. Yeah, that's me, um, and Shubrant. We have— we are building— my PhD is on video, so we are solving this with Shubrant, is coming from the hardware. So, um, yeah, I, I would like—

okay, so the plan is that we start with 10 conveyor belts, uh, with at the beginning this year, move to 40 convertibles, uh, next in couple of years. So we increase our revenue from $5 million to $23 million, and then go further up to 80, uh, convertibles to $53 million. So that's the plan. The, the, the numbers are like this, that we, we earn for each, uh, convertible $1 million, uh, over the period of 5 years. So we increase from 10 to 18. And then the, the good point is that with the hardware, we take the money right away. And then over the period, every year we take $100K for the, for the SaaS. So in 2 years, we actually make the break-even with a 13:1 LTV to CAC ratio. That's the, uh, the best part of this, our solution. Or the— so now open to questions.

Please go ahead. Thank you. Um, yeah, like, tech— from a technical perspective, maybe the microphone next time, we stick around like closer to one another and share the, the full deck, right? That's like, um, I, I guess like basics. Um, for the deck itself, I would have loved to see like both of your faces, uh, up, uh, front and see like the background. That I would have loved to see. And also images of the problem that you're solving. Um, but yeah, I'll bring Sunny on stage. Uh, Sunny, any feedback for, um, our friends here?

Hey guys, yeah, great, great job on the presentation. So steel plants lose something like a million bucks per hour, right, in downtime. What, what's your average time to detect resolve an issue?

Installation of 10 conveyor belts takes 3 to 4 months. So for 10 conveyors, and $1 million downtime per hour is for one conveyor with a major tear in a year. Normally 4 to 5 tears happen per conveyor. Got it.

Okay. And then you said the 80% in down time is predictable, right? So what does your model see that the plant operators are missing right now?

Okay, so presently, the downtime is due to the tear in the belt, due to overloading of the belt, due to something called belt sway. So the material is not positioned correctly. So there are many such issues, foreign material coming into the picture, which is too big and maybe cutting the belt, there are steel rods there. So these are the problems which can be easily resolved through the models. It's through cameras only. So it's not true that we need a lot of cameras. BELLS is running at a 6 meters per second high speed. Of course, our frame rate is very big, 67 frames per second. Sometimes we increase the frame rate, but the number of cameras is very small. It's only 3 cameras, but one conveyor belt which is almost very big, 680 meters. It's really big conveyors, it goes for miles. So you only need only 3 to cover that in few minutes.

Stefan, any questions?

Yeah, I thought, uh, I thought you have an exceptionally good business opportunity there. I really like the, the, the clarity of the problem, the value that you're bringing. Um, I, I, I'm confused by the slide that says you need $11.5 million in part to verify the ROI, because you had an earlier slide that seemed to say you've already saved $27 million for current clients. Um, so I— there's something lost in the message there, because I feel like you— there was In the beginning parts, like, man, huge ROI. And it's like, oh, but we got to prove ROI. We need money to prove ROI. And I'm kind of lost in that. So I would try to get that lined up. And I think a little more clarity around why you need $11 million versus, you know, just helping kind of the investor understand what the next steps are. And, you know, even just the— you seem totally focused on the steel industry, which makes sense, but the name has nothing to do with it. So I think there's some improvement possibilities around branding as well, just to, to, to kind of help you be successful on that. Thank you.

If you would like me to answer that, so it's— we have honestly, we have got in the funding long before, during the pandemic time. And the POCs were built in several steel factories in Pittsburgh, in many, many areas, in Newcon, in many, many areas we have built this. So they have seen the profit. So the slide that was confusing to you, Stefan, that was based on the profit that was calculated where we have deployed our POC solution. With that in mind. They are very, very conservative. It's so difficult to convince, so difficult to make the decision makers understand that this is exactly where the profit comes in. So yeah, I got the— so that slide exactly represents their downtime, their savings, and their difficulty in cutting. So all the problems that we could really resolve through the model that was install our POC.

Evan, any questions for Sarantha or comments for the team?

Yeah, thanks very much, uh, for bringing me up on stage. This is, you know, such a critical solution for manufacturing environments. Um, my question is, I'm curious if you have a data share API or if you've looked into ways to share your data, uh, in a In our case, as we model critical decision trade-offs, this is exactly the type of data that we're looking for. So we, we have many IoT sensor provider partners. Curious if, if you've looked into that sort of model.

Yeah, for the time being, we do not share this data because it's not that we do not want to. The data is very critical and confidential for every plant. They do not even allow us to take the data to the cloud. They are so conservative, they want everything should be on-premises. So that it's their, their network intranet and on-premises data. So this is where the conservativeness stands there. So getting to the sharing the data or creating open APIs, so these are very difficult at this moment. We we are trying to enter into this business, and everywhere we see that the opportunity definitely is there, and the market is very huge because they are able to see the profit. But the challenge is to explain and convince and make the decision makers into your side. So it's very, very tough, right?

Right. Hey, we're gonna bring this after close. Yeah, we see lots of potential, and then you guys like work on the pitch deck, the presentation. So thank you. Uh, let's bring the VCs on stage. Evan, thanks. Uh, Marianne, you're here. Uh, announce like who was— were our favorite, uh, today. On my side, yeah, Evan I think wins the competition, uh, for today. Just bullish on what he's building, his background. See the product-market-founder fit, uh, the mythical product- market founder fit. So I would go with Evan, generally speaking. Sunny, who was your favorite today?

I think Ilya took, took the prize for me. I was really impressed with his flow. I— what he's doing obviously is close to my heart in terms of the little ones getting education done properly. And yeah, I liked his presentation a lot.

Marianne, who was your favorite?

Uh, I missed the first one. Uh, very bullish on the last presentation. I think there's there's a very strong business there. And the one that is doing the stadium delivery, I think something there. And also, like, I'm in Silicon Valley, so chips, chips, chips. Presentation needs work, but that's my 1, 2, 3.

Steph? Yeah, I like Suthan's business, the natural baby wipes. I think there's just a huge opportunity there, um, and love the, the stick-to-itiveness that he's shown through several tough chapters. And, um, I also like— I, I— there's a bunch I couldn't quite pick up through the, the language, um, but I think the last business, Subantra's business, has got huge potential as well.

There you go. Yeah, let's bring the, the founders on stage as well. Uh, we have Sutan, we have Norbert, Albert, Axel, Sub, Evan, uh, Ilya, um, Sutan, who was your favorite today? You listened to a couple of presentations, uh, any favorite founder of yours?

Yeah, I like Evan's and, uh, Nobet's Lively, right? So just saw the, uh, the NBA final and I went with my son and it was just nightmare trying to get food for him, right? It was just crazy. So I— those two. And Evan is just from a supply chain. I used to work at SAP, so I get that aspect of it. So that's why those two connect with me.

Ilya, you checked a couple presentations. Who was your favorite today?

No, I listened to them all. It was wonderful. I, uh, I think every, every, everyone was, uh, super strong. I like the, uh, Australian idea because it's both the delivery of Australians today. Kind of methodology. Evan is obviously a very sophisticated player. Uh, the chips, just so over my head. But, but like, it's definitely— there's definitely so much competence there. Just everyone was super strong. Evan, any favorites of yours today?

Well, uh, Axel and Subranta, I understand, you know, what you're working on. I thought I thought it was really interesting and yeah, definitely keen to connect and keep in touch because I think we might be able to potentially collaborate. Subranta, you've—

both of you, you've listened to a couple of presentations, the shows. Who was your favorite?

I would say I really like Axel's presentation, Abacus Semiconductor, because I am a software something doesn't make sense until and unless I get exact computation-intensive hardware. Yeah, but yeah, that was, that was extremely nice and I like— so every presentation was much, much better than mine. I really think that's my thought process. And we work on that, don't, don't worry.

Axel, any favorite today?

Not really.

Any favorites. I think all of the presented businesses had their merits. I think the issue is that presenting something complex is always more complicated than something easy and simple.

So, you know, the— I think the businesses all are pretty strong, so I don't have an issue with the businesses underlying it. And I, I agree, my pitch needs redo. That's, I think, obvious. It's, it's still too complex despite the fact we took out—

didn't we do a pitch for you? Where was that? Pardon me? Didn't we do a better pitch deck for you, me and Sunny? Where, where was that?

So I had a look at that and it's missing a few things that were critical. The efficiency isn't there at all. And then, you know, the other quadrant that I added back in that needed to be there, and that wasn't in that either. So I think Sunny's approach was certainly better than what we had in the very beginning. And it's— believe it or not, the one that I presented today had already 90% less tech in it than the one we started out with.

So yeah, which is great.

Norbert, any favorites of yours today?

Uh, Suzanne and Ilya, I really love your business businesses. I mean, my kids are a little bit older, uh, you know, it's, uh, I, I know the dance, right? It's, it's something that's very, very close to home. Um, and, uh, yeah, I mean, I thought your presentation was just tight. Um, uh, so I really enjoyed that. But I mean, you know, I think all the, all the presentations questions had their merit. Um, obviously I love my deck as well, so, uh, I love it.

Uh, well guys, thanks for connecting today. That was, uh, the 5th edition of, uh, Spark Tank. Uh, you guys, uh, did a great job, I think, like all the founders, all the VCs. People can find us at, uh,