Media Week NYC theCUBE + NYSE Wired | Kirk McKeown, Carbon Arc

Clips
More from Media Week NYC theCUBE + NYSE Wired

Kirk McKeown

Co-Founder

Carbon Arc

play_circle_outline Carbon Arc's marketplace solving data structure problem by making data consumable for customers.

play_circle_outline Challenges in data include governance, proprietary vs. open data, data provenance, and quality.

play_circle_outline Carbon Arc helping customers improve hit rate, make better decisions, and inform business opportunities.

play_circle_outline Faster and more accessible data consumption with Carbon Arc compared to traditional models.

Info
Transcript

Kirk McKeown, Carbon Arc

Kirk McKeown

Co-Founder Carbon Arc

Covering the New York Stock Exchange, TheCUBE features Kirk McKeown, CEO of Carbon Arc. The company aims to solve data structure issues by creating an insight marketplace where data suppliers can bring their assets to be structured for customer consumption. This approach aims to democratize the data stack and aid businesses in decision-making. Carbon Arc offers a pay-as-you-go model to improve conversion rates and access data quickly. By providing a consumption-based exchange, the company simplifies the process of accessing structured data through APIs and da... Read more

explore Keep Exploring

What problem is Carbon Arc stepping into the middle of and trying to solve in the AI economy? add

What are some of the challenges and goals related to data provenance, payment for intellectual property, and implementing a markets-based approach in the data market structure? add

What is the company's approach to meeting customers at different levels of their digital transformation journey and how are they addressing the challenges of data monopolies in the marketplace? add

What is the process like for buying and selling data sets in comparison to trading equities in the 1930s? add

bolt Powered by CUBE AI

Kirk McKeown, Carbon Arc

search

>> Hello, welcome to theCUBE's coverage here in the New York Stock Exchange. I'm John Furrier, your host of theCUBE. We are here on our set. We're dropping in for a media day all three weeks here in the New York Stock Exchange. We're going to be looking at all the top trends around climate week, climate tech, I'll say the UN's in town, and we're going to actually unpack the global nature. Of course, topic is going to be AI. And our next guest, Kirk McKeown, co-founder and CE of Carbon Arc. Kirk, great to see you. Thanks for coming in.

Kirk McKeown

>> Hey, thanks for having me.

>> So got nice backdrop here. We're on the floor of the Stock Exchange. This is where business and technology are intersecting. It's a busy week. The top story in the world is the impact of how people are live in their lives and running their businesses operationally. But then the generative AI wave is here and you start to see that nexus forming, where you're seeing a transformation across all theaters, every industry, every government, policy, all aspects of it, the data is changing everything. The stack is looking like it's evolving quickly into being data-driven. You're at the middle of it. What's going on with ? Give us a quick . First take a minute to explain what you guys do.

Kirk McKeown

>> Sure. So my background is I spent 20 years in investment management on the buy side on Wall Street, and I built research systems at those shops, sitting in that seat. I was buying a lot of data and the thing that you learn buying a lot of data on Wall Street is that it's easier to buy a house or a car than it is to buy a data set in the United States. When you start thinking about the supply-demand dynamic for data sets, they're really expensive, really hard to work with and all that fun stuff. Carbon Arc is stepping into the middle of that to solve that problem. If you think about the AI economy, there are three core components we break them up into. The models, the chips and the data structure, and we're attacking the data structure problem. We're basically fundamentally building an insight marketplace where data suppliers can bring their data assets to Carbon Arc, get them structured in a way that's consumable for our consumers, and our customers come to the marketplace to buy data.

>> You know what's interesting is been covering this big data wave for 15 years since SiliconANGLE was founded, when I started the company with Dave Vellante. Back then, Hadoop was the big story if you remember those days and it was hard to set up. But now with cloud technology and now with the generative AI, you now have scalable infrastructure. On the semiconductor side, look no further than NVIDIA. Look at the stock was up yesterday, again, the market hit an all-time high. Semiconductors at the silicon level, custom silicon, and then at the data layer, you've got open data formats, you have now intelligent applications being built. So you got this data operations going on and the ability to acquire data, now it becomes a competitive advantage. And if you look at all the aspects of the top apps, their programming down to the kernel level, systems level, at the same time, making as much data available to the AI systems as much as possible. The more data, the more intelligent. So okay, sounds easy. I buy the logic, but it's hard. Okay, so take us through what some of the challenges are today because you've got governance. Is it proprietary data? Is it open data? Is it clean data? Has it been infected with malware? Has it been infected with prompt injections? Has it been context poisoned? I mean there is a slew of challenges ...

Kirk McKeown

>> That's right.

>> In data. What are some of the things that you see?

Kirk McKeown

>> Well, the challenges are all the ones you've named. There's also data provenance. There's people getting paid for their IP, so data asset owners being commensurally paid for consumption of their data. What we are looking to do is bring markets-based approach and financial system-first principles to the data market structure. We look at the GPT moment in AI as the Black-Scholes moment on Wall Street. And so fundamentally, we think what's going to happen on Main Street and corporate space over the next five to 10 years is what happened on Wall Street over the last 40 to 50.

>> It's interesting you mentioned Black-Scholes. That was a critical moment. The other one in Wall Street was high frequency trading, getting an edge on speed, move the packet speed , get closer to the trade, make more money. Now we're seeing high frequency insights, you're seeing high frequency productivity. And I think one of the things that's coming out, and I know you guys, I saw it on your website, you guys believe in this, is that productivity and human insight's huge.

Kirk McKeown

>> That's right.

>> This is a data advantage. How are your customers looking at this? Because you got to get the data. Are they mindful of the productivity angle or is that still an operational challenge? Where are your customers at?

Kirk McKeown

>> So customers are across the gamut. At the top of our customer stack, we have AI-native companies that are coming into our stack to pull data down through the pipe at meaningful levels, to power models and train models all the way down to people who've never used data to make decisions before. And that digital transformation journey, we're meeting people where they are. At a fundamental level, you start with a data set on the left side of the ledger and you have to make a decision on the right side. So one of the things we always said on Wall Street is it doesn't matter the insight, you still have to trade the stock. You still have to take risk. And so all we're trying to do is build into the use cases and modular decisions that every company needs to make to inform their decisions with customers and their decisions with suppliers and really understand that demand management framework.

>> I want to unpack the marketplace aspect of data because everyone who watches theCUBE knows I'm a huge fan of marketplaces. There's so much efficiency in it, there's more optionality, everything really works well in marketplaces. Let's talk about Carbon Arc and the origination story, where you guys are at. Take us through the company history and then what's the status funding, employees, customers?

Kirk McKeown

>> Sure. So sitting on Wall Street, started to look at and understand the data market structure and the imbalances in data. And it was natural monopolies, data monopolies, Nielsen and IRI and CoStar, Facebook, Google, Bloomberg, all of these created massive silos, so data didn't flow. It was really expensive, really hard to work with. You need engineers, you need scientists to make sense of it. At the end of the day, we wanted to create a marketplace where the transaction was the insight because that's the payload to make the decision closer to the transaction, which is that final decision. Started Carbon Arc in March of 2021 and started to buy data, to structure it the right way, to deliver, through API endpoints and dashboards, all of the different inputs that a consumer, whether it's a big three consulting firm or Wall Street hedge fund or a sports organization, to make better decisions around when they should have their next event or who should be at their next concert or whether they should remodel stores. So we've raised about $50 million to date. We've got about 30 customers. That's across the gamut from Wall Street to Main Street. We've been broad in our verticals. We have consumer products companies, we have sports organizations, sports teams, live event managers, Wall Street firms, across the gamut. We've got about 30 employees and we've got 60 suppliers on the left side of the ledger delivering their data into our exchange and they're getting paid on consumption and we're getting paid on consumption. So we've created a two-sided consumption-based exchange.

>> I love that. And take us through the data structure problem. What are you solving there? Because again, data is complicated. When you start thinking about lineage, where has it come from? The supply chain aspect of it. I like how you got this with the end points with API's clearly the way to go. What does it mean by you helped structure the data? Take us through that problem and what you guys did.

Kirk McKeown

>> So the data structure problem is actually, at core, what provides the scaffolding for the modular frameworks and making decisions. At the end of the day, we live in a world and believe that every company in the world is trying to figure out where their inventory is, how do they move it and how do they sell it. And so we've modularized all of those different funnels and frameworks and boiled down those use cases against entity structure and event structure. So our entity structure is really based around companies, brands, people and locations. We pivot around those points in graph and we deliver against an event structure because we believe that you need an event to make a decision. And where Wall Street and Main Street collide is it that mirror of company decisions. Management teams are trying to improve their revenues or cut their costs and Wall Street's trying to figure out if they're doing a good job at it.

>> Why are companies buying from you? Is it because they want more insights? Are they solving the data problem? Are they getting more access to data? What's the main challenge?

Kirk McKeown

>> It depends. Some of it's just data access. So we have all this data structured and you pay for what you consume. So you can actually digitally transform against your seat in Carbon Arc because you now have all this ability to work with data to answer the questions that you need answered around targeting or retaining customers, around visibility into your supply chain. And so some of it's data availability, a lot of it's price. At the end of the day, people need the data that they need, but they don't need all the data that comes with buying something from a large data provider. And so the data providers, like us, because we create distribution into a part of the market that they have never captured before, and our consumers are coming to market because they could buy data as an OPEX, not a CAPEX.

>> Yeah. And they get more access. So it's not controlled by the gatekeepers that were controlling the data sets in the past.

Kirk McKeown

>> That's exactly right.

>> And it's mostly public data. On your side in the marketplace, is it all public data?

Kirk McKeown

>> It's commercial providers. We have open source data, but we also have closed source, where we partner with commercial providers who are coming to market and selling their data into in our exchange.

>> Give an example of a customer use case.

Kirk McKeown

>> A customer use case. One of our partners is a professional sports team. They own their stadium and on off nights, they have to figure out who's going to play it. So they've used our data stack to figure out which concerts they should be playing in their market and how they should think about curating food and beverage based on those folks who like that specific act. That's one use case. Another use case is thinking about which brands they want to have partners at different events for a large professional sports league. And so really creating improved conversion rates around targeting, acquiring, upselling, retaining. So living in that CLTV equation.

>> People watching right now might be interested, why should they buy from you? What's burning in their world that would change in working with Carbon Arc? What's the key things? Is it market intelligence? Is it more time to market on the value of their assets? What's the main thing that people would be, "Hey, I should be leaning into this?"

Kirk McKeown

>> It's improving their conversion rate. And at the end of the day, anytime you're buying data as a data buyer, you have to inform one of three things, creating more opportunities for your business, improving your hit rate or making bigger bets, whether you're trading on Wall Street or you are managing a wholesale consumer goods business. Those are the three things. We think we live in a world where we're helping people improve their hit rate.

>> What's interesting is that you guys are democratizing the data set business by creating this marketplace. The pay as you go is great with cloud consumption model. You see that APIs make it accessible for applications. What's next? What's the main innovation that you guys are doing? Is it availability of the data sets? Is there any secret sauce that you guys have built into the system?

Kirk McKeown

>> Well the system design is all open source and we've forked a number of open source technologies and built a framework that works and we're on the knowledge on top for the data structure and entity structure that powers modular decision making. Where it goes is not dissimilar to what happened on Wall Street over the last 50 years. So the way we're building everything is around driving down on a per consumption basis and really evolving the different products that people can consume in their models to make better decisions. We think that you can factor Taylor Swift. She has momentum, resonance and persistence and those are factor frameworks that follow the CAPM regime.

>> Take me through how you see the market in terms of consumption, on the consumption side. Okay, you got a marketplace, you got a lot of data sets. So I'm a user. Consuming that is going to be what I really care about. Obviously the conversion for sales, my business, okay, checks all the boxes. How does the old model compare to what you guys are doing that's innovative? Is it speed? Is it consumability? Without you, what would be the choice of the customer? Setting stuff up? Big contracts?

Kirk McKeown

>> It's the old setup. The way I always say it is that data sets trade like 1930s equities. You need a big bag of cash, you need to be a market operator and you need to show up at one of the 13 exchanges in the United States to walk away with a block of stock. Data buying and selling is very similar now. It's a long lead time and sales cycle from a legal perspective and a contracting perspective. You need to hire a couple of engineers and a scientist to make sense of it. And then it's trying to figure out that ROI attribution on the other side of that CapEx build. It's time, it's money and it's a capital outlay. Here, you just get stood up, we're your counterparty. We've vetted everything against your terms of use and the data provider's terms of use. You plug into the APIs of the dashboards and you can start making decisions in a week.

>> Yeah, I mean that's the key thing. That's why cloud is great, changed the world over a decade ago because the ability to stand something up, whether it's a data center, cloud comes in. Data sets, huge money, huge contracts, set up the plumbing, all the engineering involved and then the compliance. I mean where's that data coming from? What's the terms of service? All those things create huge lag times and most people will just abandon it and they're like, "Okay, I'm done. Let's tap out." That's the old way.

Kirk McKeown

>> And that's what you've see with the corporate space, specifically. Big tech being data-native and web-native, they have a critical advantage in digital transformation. But when you look at the bulk of corporate America, there's been a dearth of data and they've been focused on engaging with benchmark-type data and now there's an opportunity to expand on that. We're seeing a lot of interest in that.

>> So take me through the journey as an entrepreneur. Just started the company a few years ago, you're in it now. What's that been like for you? You guys got some good financing behind you, you got some good cash, got some great customers, developed a horizontal solution, every vertical needs data. Take us through your journey and how you see the future.

Kirk McKeown

>> Yeah. So I spent 20 years on the buy side on Wall Street and I built businesses inside of large hedge funds. And in doing that, I would say that building a business outside has been humbling. It's been like building a business on a Boston Whaler in the North Atlantic. Before, I was building on aircraft carriers in Lake Michigan. So having a mandate from senior folks at these firms and then having captive customer bases inside of those organizations. I didn't appreciate the double cold start problem that we tackled the Carbon Arc and that's been super humbling and really challenging at different times. But really having a big vision and getting up to every day with a mission-driven approach to democratization of insight has been a huge tailwind in keeping the ball going. Where I think the world is going? I actually think, in 10 years, you're going to see these types of data inputs traded, not stocks. I actually think they're closer to ground truth and model to model decision making. Models will be coming to market every day. Agents will be buying data and selling data and I think exchanges are natural, centralized frameworks that will exist in this coming world.

>> I think that's a great vision. I mean one of the things, I think why I support that is because if you look at where the innovation is, it's going as close to the value as it possibly can.

Kirk McKeown

>> That's right.

>> In this case it's the inputs to the trades. Now you've got agentic systems coming online. So the first wave is get the hardware there. Nvidia and others are leading the way in terms of democratizing supercomputing.

Kirk McKeown

>> That's right.

>> Okay, I got tons of resource, now you're seeing more and more, that's why their stock's trading so high. Then the data layer is going to emerge. Okay, making sure it's agile, programmable, and then now data inputs come in. The applications will be coding into the data sets.

Kirk McKeown

>> That's right.

>> So as people move down closer to the truth, that's happening everywhere. , my friend in Silicon Valley says, I won't say the name of the company, but they're a big AI company doing search. They're programming to the kernel level for a speed advantage because once they hit the prompts, they go into the neural network, they're already pre-optimizing the next prompt, not just giving the generated results. So you're already starting to see the intelligence and the performance of the systems working in that direction.

Kirk McKeown

>> That's right.

>> This is a huge thing. And then so the next question I want to ask you, because I think this is a thread that tees up the next chapter of the journey that we're all living, is we published on theCUBEResearch, a power law. A year and a half ago is the first ones that put out a power law, the LLMs at the top of the power law, OpenAI and the big guys. And then we saw the specialty small language models emerging and we were first to see that. Now private AI is all the rage. So you start to see the notion of data sets that are either proprietary or pretty well-trained and have enough context, but they're small, they're not big. Models are starting to talk to each other. So I'm imagining that, in your world, data sets will have intelligence about each other and soon will be programmable. What's your reaction to that?

Kirk McKeown

>> I think that makes a lot of sense. One of the things we're thinking about is, as more and different data sets proliferate, you're going to start to see some blurring of lines because it'll get down to the transaction level and the cohort level and the entity structure will emerge. And it's important to have attribution across so that people get paid commensurate for their IP. That's one of the big challenges that we've seen in the data structure is that, while there's been this huge movement in large language models over the last five years, there has been a big question around IP ownership and whether it's Conde Nast or New York Times, they have had a focus on being taken care of and paid for commensurate for their IP. And so I do think those things need to get sorted out. Think of exchange as the best way.

>> Yeah. And we're going to keep an eye on that. Obviously governance is a big deal. Quick final question, what's your experience with governance been? Because you're shortening the ability, lowering the barriers and shortening the time it takes to get these data sets. One of the problems that you guys solve is all that paperwork, all that hassle that's involved in, okay I have to sign a contract with this data set. Is it the right data set? What's your experience been with governance of data?

Kirk McKeown

>> It's moving fast and in terms of transformation, and providers are getting more thoughtful about how they want to structure their contracts. What's really interesting is we actually look at a supplier coming onto our exchange almost like an IPO. They have to get ready for the IPO and so they come on. We have standard terms of use that we've worked out with so many folks that we're actually engaging with the conversation to get them commercially ready to go because there might be 2,000 commercially available data sets out there today. It's going to 200,000 over the next three years.

>> So it's like cloud. You got to do a little bit of work on the front end. You can get in easily, do a little bit of work, but once you're in, you're building. You're getting value out of it quickly.

Kirk McKeown

>> That's exactly right. Yeah. We go from a signed contract to ingest to structured in like three weeks.

>> That's awesome, Kirk, great to have you on theCUBE here, as we drop in and start kicking off our studio here on the East Coast. I call it our access point to our network. Great to have you here. Just to end the segment, give a quick plug for the company Carbon Arc, what you guys are looking for. You're hiring. Customers are watching, people are watching. Give a plug for the company.

Kirk McKeown

>> Yeah. So we're open for business. We're trying to democratize the insight stack to help every business out there make better decisions in a way that is scalable and additive to their business. Our data plus your data equals a better outcome. We're hiring engineers, we're in the market for go-to-market specialists and customer success and we think that this could be the transformative company and movement in the data stack and data structure market.

>> All right, Kirk, great job.

Kirk McKeown

>> Thanks.

>> Okay. I'm John Furrier, host of theCUBE. Kirk McKeown, co-founder and CEO of Carbon Arc. Check it out. Data marketplaces are hot. They're only going to get better. They're going to shift towards the sources. If you've got data, the world is going to be great with generative AI. I'm John Furrier, here on the New York Stock Exchange, here at our new home on the East Coast. Thanks for watching.