theCUBE + NYSE Wired: Mixture of Experts Series | Kim Crawford Goodman, Smarsh

Clips
More from theCUBE + NYSE Wired: Mixture of Experts Series

Kim Crawford Goodman

CEO

Smarsh

play_circle_outline Rainbow Room, Rockefeller Center Customer Event: Multi-Channel Platform Updates Including Email, Text, Voice, Video, AI (Bedrock, Claude)

play_circle_outline Smarsh using AI to shift from comms compliance to communications intelligence.

play_circle_outline Need for comprehensive capture due to new apps and AI agents (WhatsApp, Instinct).

play_circle_outline AI surveillance reduces false positives and highlights real compliance risks.

play_circle_outline Communications Data Beyond Compliance: Expanding TAM into Government, Healthcare, Pharma, CRM with Discovery and AskSmarsh Agents

Info

Kim Crawford Goodman, Smarsh

Kim Crawford Goodman

CEO Smarsh

search

(INTRO)

Gemma Allen

>> Welcome back to theCUBE Studio here at the New York Stock Exchange. I'm Gemma Allen, co-host of NYSE Wired's Mixture of Experts. And now we're going to have a conversation about how enterprise communications have exploded from email and phone to Slack, WhatsApp, Teams, and now agents. For regulated companies, that creates a massive challenge. How do you capture, govern, and make sense of all that communication? Smarsh has built its business around exactly that. Now it's using AI to move from comms compliance to what it calls communications intelligence. Joining me to help understand exactly what that means is Kim Crawford Goodman, CEO of Smarsh. Welcome, Kim.

Kim Crawford Goodman

>> Thank you very much. It's a delight to be here, Gemma.

Gemma Allen

>> So you were here on the show in April, and I know you're back here in New York because you guys have a customer event here at the magnificent Rainbow Room at the Rockefeller Center tomorrow.

Kim Crawford Goodman

>> Yes.

Gemma Allen

>> But I am sure an awful lot has happened since you got together with your customers last year and even since you were on the show with John Furrier. Had a quick recap on those notes and I was like, wow, so much change in such a short space of time.

Kim Crawford Goodman

>> Yes.

Gemma Allen

>> So maybe just to start, bring us up to speed. What has been happening for you and the team at Smarsh in this last 6 to 12 months?

Kim Crawford Goodman

>> Well, as you know, Gemma, at Smarsh, we are the largest company in the space of communications, data and intelligence. We are, as best we can tell, more than 2 times the size of our nearest competitor. That means that our R&D budgets are multiples of many of our competitors. And what is happening now is the world is moving faster than it ever has in technology, in compliance, in risk, and of course with AI. So in just 6 short months, there are many things that we have done. First and foremost, we've lit up our entire platform with AI on the side of capturing the communications data, which you must do. You must capture all the data. You must capture it in a good and rich format. We've lit that up with AI. So now we don't just capture email, text, voice, video, all the things you would've known of, say, a year ago. Now we're also capturing Amazon Bedrock and Google Workspace, and of course, Claude for Enterprise, as well as any of the AI tools that our clients need because they need to be able to capture and monitor what's happening with their AI in order to keep it safe.

Gemma Allen

>> Well.

Kim Crawford Goodman

>> We've lit up our fundamental platform. We've made our platform fully comprehensive because many times our customers have pockets and silos of data and capability that speaks to their communications data and speaks to how they use it for compliance. We're pulling all of that into one very capable integrated platform and giving people more access to it. So we've lit up the fundamental platform. We're making that platform more useful with a whole suite of agents that I'm sure we'll talk about some because the AI is not just crawling the data now looking for bad behavior. Agents can work in the data and accomplish a lot on our behalf, on behalf of our customers. And last but certainly not least, we've lit up the platform through access. Of course, like every technology, we've had access for many years now with APIs. But since we were here in April, We've announced our MCP server, which allows people's agents to access actually our platform, the data and the agents on our platform. So there's so much that's happened in just 6 months.

Gemma Allen

>> Well, you're certainly meeting the market moment, that's for sure. There's so much to unpack there. I really want to talk about MCP and I want to talk about Agentic. But first, I want to start with something quite basic, right? And that is that in a short space of time, what it means to communicate and who's communicating has changed pretty dramatically. Last week we saw the rollout of Instinct. Are you familiar with this invite-only viral app that folks are downloading? You can use it through WhatsApp or text message and ask it to do all sorts of things for you. Book your holidays, talk to you, advise somebody at work you'll be late for a meeting. Right. And it's taken off so quickly, I almost couldn't believe it myself when I read the news piece this morning. So we know who's communicating with whom and when, it's become a whole new opportunity and conundrum. From your perspective, as all of these comms channels are exploding, what are your thoughts around how things are being governed at an enterprise level? What sorts of risks and compliance challenges have you guys seen?

Kim Crawford Goodman

>> Yeah, this is what you're articulating is one of the age-old challenges of communications and of compliance and of intelligence. That just continues to roar forward. What I mean by that is communications capabilities are always expanding. If we had been talking just a year or two ago, we would have been talking about the billions of dollars of fines that clients had received because all of the mobile apps had come to bear and you had investors communicating with their advisor or you had traders communicating with each other. On things like WhatsApp and WeChat and Signal, and it was not properly captured. And that resulted in a compliance failure and literally billions of dollars of fines. Now we see the explosion of AI capabilities. As I just mentioned at Smarsh, we are always first to market to capture what is actually going into this AI and what's coming out of it. And we know that there will be a need for that in the future. We talk to our clients all the time and they are always describing whether it's in Slack or something else that new technologies have just added a chat capability to that technology. Whenever these things are happening, it must be captured. It must be captured in rich and good format and it must be available for the intelligent capabilities of people like Smarsh to crawl. And now we're looking not just for risks and bad things, we're also looking for opportunity. At Smarsh, we describe that what we do more than anything is we preserve and protect the truth. And in this day and age, you must know this better than anyone as a reporter, what is actually true has— right, has— it's changed how you find what is truly truthful. And there's a lot in the atmosphere that is actually asserted but is not necessarily truthful. So it makes the capture, the use, the intelligent evaluation of this communications data particularly by regulated companies, by government industries, by financial companies, it makes it even more important and it's even more important for it to be completely comprehensive.

Gemma Allen

>> So if we think about the model a year or two ago, a year or two ago, within any enterprise, right, you need to capture data, you need to archive that data, you need to be able to search and discover that data, and then you need to surveil that data, right? Like that is absolutely essential to what needs to happen. In an agentic world, that needs to happen like a million X, really, right? Because we don't know how many instances of new data are going to be created and stored in any one environment on any given day, especially in a world where AI agents are actually accessing systems on behalf of humans. So when you guys go to build out this technology stack around that problem, where was the key focus first? Was it on, a better version of Capture? Was it on the surveillance end from the perspective of cyber? How has it kind of changed what you guys are building towards from a product perspective and where do you see this headed?

Kim Crawford Goodman

>> Yes, it's an excellent question. And what you described is the fundamental process for compliance. And it's also— you're also touching on why Smarsh and our scale is so advantageous to our clients. As in many industries, when you have larger scale, you can invest more, you can enrich more, and therefore you can add more value. So we are investing both in the compliance workflow that you just went through, but also how do you take the value of this communications data beyond compliance because it is so rich. On the compliance side that you just described of capturing, archiving, and surveilling, we are investing across all three. The capture has to get broader and broader. And we've introduced, as I mentioned, capture across all kinds of relevant AI platforms as we continue to chase every communications vehicle. We bought a company early last year that does voice, because what we see is that to really utilize this data properly in the future, peo— our clients are gonna want all communications together, not just the email and the text and the WhatsApp in one space, but they're going to want all of the voice and video all there together to be searchable. But we have invested in the surveillance because what happens in the business communications is that many times people are surveilling, they're looking for risk, but it's like looking for a needle in the haystack. And so it is very helpful with the right AI technologies to actually look for real risk It means that you are eliminating things called false positives. It means that you're making people's fundamental operation and compliance far more efficient because your false positives go down by 70, 80%. And what the resources are really being used for is to analyze and find real risk. And so we are investing so our clients can utilize their resources on real risk and not things that are false. But then we are investing outside of compliance as well and creating capabilities for this data, which is so rich, to be used in other areas. A great example of that is we have an agent, we have a whole suite of agents, but one of them is called Discovery Agent. It is for investigations. It's for legal work so that if there's some issue or dispute, say between Gemma and Kim, Instead of just doing a dump of every single piece of data from you and me, anything we've ever said, period, or to each other, now what can happen is the agent can go in there, the agent can crawl all kinds of communications, the agent can see the context. I've seen a demo of this and the agent can actually create a context graph of every time we talked to each other, but then every single person you talked to about me or I talked to about you, and you can see kind of a graph of how This investigation needs to proceed based on how communications expanded.

Gemma Allen

>> Wow.

Kim Crawford Goodman

>> So the investment we've made is tremendous on the compliance side to be at the very forefront, to work in a partnered way with our customers on our Smarsh Communications Intelligence Platform that is lit up by AI, but also to invest in a suite of agents that can help them use this technology and use this data to help their enterprise more broadly.

Gemma Allen

>> So I want to get on to MCP because I truly, when I looked at your business, I thought that I think that is a very smart move, right? You have to roll with the direction that this industry is headed. And we know that folks are going to design their own agents. They are going to want to call systems. So allowing them in, I think, is actually truthfully a great idea. Talk a little bit about what it means from your perspective for your TAM. You and I will have agents that are external coming in looking through. And your data, I'm sure, is like a huge part of the moat. It's a sticky nature of this business, right? It's a powerful piece that you bring. How did you think about that? When you guys are weighing up that decision to go forward with this MoE model, especially in the world of frontier models and LLMs. Talk me through what sorts of ways in which you guys debated that and where you landed on it.

Kim Crawford Goodman

>> Well, we're actually going to be talking to our customers directly about this tomorrow at our customer event that you mentioned. What we see very clearly, because we work across the industry in financial services as an example, we work with 18 of the 20 largest financial institutions outside of mainland China. What we see across the industry is that there are going to be all kinds of models used. So in the press, you all do talk quite a lot about the large language models, and they are very capable and they're a very important innovation. But what we see already happening with our clients and will happen more is they will— so they have access to and they will select from a variety of models. The large language models will be out there. But there are already open source models that depending on who you are and where you are, people are choosing from. But there are also a multitude of small language models. And we at Smarsh, because we've been involved in AI even before it was AI, they called it machine learning. So we've been involved with this for 10 years. We have actually refined a small language model all around communications data and what it can do on behalf of our clients. And because it is a small language model, it's very fit for purpose, both in terms of how good it is, but also how efficient it is from a cost standpoint. So what we see at Smarsh is people will develop and utilize all kinds of agents from different sources, and they will do that on the back of different types of models. So we are developing native agents to our platform that people can use. We are doing that on the back of our small language model. However, we see that our clients will develop their own agents, our clients will want to bring those agents to integrate with and utilize the quality data that we have because we haven't talked about it. But one of the things that's true in the world is not all data is the same. Much communications data is messy. Much communications data doesn't properly ascribe Who actually sent what and when did they do it? Our data is foolproof and pristine because it has to be used for legal cases, for regulatory, et cetera. So our clients will want to develop some agents. They'll want to come and use our small language model. They'll want to interact with our data. Their agents will eventually talk to our agents, and that has to be done in a sophisticated way through MCP servers. So that's why we were very keen to invest in this, very keen to develop it and bring— and be the first to bring it to market.

Gemma Allen

>> Like I said, I think it's a great idea because you need to meet customers where they're at. And this is a confusing moment in technology. A lot of folks are making buying decisions, not really knowing what 5 years out looks like. So understanding how you can continue to be part of their journey is huge. I want to ask you about the competitive landscape that we're in. we ask all folks about this.

Kim Crawford Goodman

>> Yes.

Gemma Allen

>> When I think about the surveillance element of your business as a good example, we've talked to a lot of folks in cyber. A lot of money being spent in R&D in that space. And sometimes I think about it from the perspective of convergence, right? We often debate with CEOs and founders on this show. Will we see a mass convergence? Will we see a world where, to quote an unnamed Frontier Lab leader, one orchestrator will do everything? How do you guys think about the competitive dynamics that you're in? You have companies like Microsoft and I know there's big players, smaller players who I'm sure are also coming for this long-tail opportunity. Right. I'd love to understand from you, Kim, how do you think about that and what does the next 2 to 5 year priority look like from that perspective?

Kim Crawford Goodman

>> You know, we have a range of competitors that we have a great deal of respect for that are formidable. I used to work for Michael Dell. And one of his philosophies at that time was it's good to have good competition. Yeah, because it keeps you on your toes and it makes you excellent. So, yes, we have competitors that are traditional and focused in our space, although a fraction of our size. We have competitors that are kind of new, funded, brand new and somewhat AI-oriented that are even smaller in size. But that, as you mentioned with the new applications, that can change very quickly. But what they don't have is the depth of capability of all the security that's needed, all of the interfaces that are needed. There's a whole— you almost can think about it as an iceberg. And there are very beautiful things at the top, demos that people can do, but there's so much capability that you have to have underneath to have the data right, to have the scalability, to have the global capability, to have the repeatability. And then we do have, large companies. Obviously Microsoft is a major communication provider. What we have found is the needs in this space for regulated companies go so deep that many times when people are either brand new or very broad, they don't satisfy the needs as deeply as we do. You asked me about our investment areas. We don't have 50, 100, 1,000 investment areas. We are always making the usefulness of this communications data better, and that is our absolute focus. And that's what keeps us ahead of competition.

Gemma Allen

>> Tell me about new opportunities you're seeing on the market. I know you guys have a huge lion's share of the financial services industry. Where else are you seeing new needs arising because of this agentic world we're all in?

Kim Crawford Goodman

>> Yeah, I would say first and foremost, we're seeing an expansion of people needing this quality of communications data. For example, in places like government. Government has a Freedom of Information Act. If they close the elementary school down the street from you, you have a right to go to that government entity and ask for every email, every, hearing, every document that existed that justified the closing of the elementary school where you thought you were going to send your son or daughter. Health care— we know that health care has all kinds of laws of appropriate privacy and appropriate decision-making that affects the life and death of all of us as individuals. So we see an expansion of the market from a TAM standpoint for the quality of this information. We also see an expansion of the market for the use of the information. So I described, for example, the Discovery Agent, which is a legal use that is not restricted to people who are regulated. We have a whole suite of agents that are— some are introduced and others that are coming on this communications data. A perfect example is we have an agent called AskSmarsh. This allows employees, just like you can query the internet in natural language methods, you can now query your own communications data in very natural language methods. And because this is private and pristine information, it's not accessible by the large language models. The way it's accessible is by the AskSmarsh agent. We see use cases for driving revenue growth over time from a CRM standpoint, because sitting in this information is everything that a customer asked for, how employees responded. And you can combine that with other datasets to see what actually enabled things to be sold versus not. So we see opportunity on the communications compliance side from new industries that have requirements like government and healthcare, pharma and others. But we also see tremendous opportunity across multiple industries for other use cases well beyond regulation and compliance.

Gemma Allen

>> Well, I think if you can offer enterprises and all customers access to structured data that's protected without the risks of data retention and all of those things, you certainly have a model that's needed in this moment. We're in right now. So, Kim, last question for you. I know you're here in New York. You're meeting some customers tomorrow at the Rockefeller Center. I guess what's your, what's your main message for this week? What do you want everyone to leave that room knowing tomorrow on behalf of you and the Smarsh exec?

Kim Crawford Goodman

>> My main message for the week is that your communications data is tremendously important and Smarsh is making it more and more intelligent.

Gemma Allen

>> I love that.

Kim Crawford Goodman

>> For you to use in ways that are more efficient and effective. If you already use it for compliance regulation, But it is also for use of things that are greatly expansive, whether that be legal, whether that be cyber-related, whether that be searching, whether that be gaining more revenue. So the opportunity— we are in a moment where the risks of being left behind are huge. Every entity, every professional has a risk of being left behind, but our clients don't need to worry about that because when they partner with us, we are deeply invested in keeping them at the forefront and giving them real value.

Gemma Allen

>> Well, I love it. It is certainly a moment we need to think as intelligently and as effectively as we can. So, Kim, thank you so much for joining us on NYSE Wired.

Kim Crawford Goodman

>> Thank you very much for having me.

Gemma Allen

>> I'm Gemma Allen here at theCUBE Studio at the New York Stock Exchange. This is NYSE Wired: Mixture of Experts. Thanks for watching.

Kim Crawford Goodman, Smarsh

search

(INTRO)

Gemma Allen

>> Welcome back to theCUBE Studio here at the New York Stock Exchange. I'm Gemma Allen, co-host of NYSE Wired's Mixture of Experts. And now we're going to have a conversation about how enterprise communications have exploded from email and phone to Slack, WhatsApp, Teams, and now agents. For regulated companies, that creates a massive challenge. How do you capture, govern, and make sense of all that communication? Smarsh has built its business around exactly that. Now it's using AI to move from comms compliance to what it calls communications intelligence. Joining me to help understand exactly what that means is Kim Crawford Goodman, CEO of Smarsh. Welcome, Kim.

Kim Crawford Goodman

>> Thank you very much. It's a delight to be here, Gemma.

Gemma Allen

>> So you were here on the show in April, and I know you're back here in New York because you guys have a customer event here at the magnificent Rainbow Room at the Rockefeller Center tomorrow.

Kim Crawford Goodman

>> Yes.

Gemma Allen

>> But I am sure an awful lot has happened since you got together with your customers last year and even since you were on the show with John Furrier. Had a quick recap on those notes and I was like, wow, so much change in such a short space of time.

Kim Crawford Goodman

>> Yes.

Gemma Allen

>> So maybe just to start, bring us up to speed. What has been happening for you and the team at Smarsh in this last 6 to 12 months?

Kim Crawford Goodman

>> Well, as you know, Gemma, at Smarsh, we are the largest company in the space of communications, data and intelligence. We are, as best we can tell, more than 2 times the size of our nearest competitor. That means that our R&D budgets are multiples of many of our competitors. And what is happening now is the world is moving faster than it ever has in technology, in compliance, in risk, and of course with AI. So in just 6 short months, there are many things that we have done. First and foremost, we've lit up our entire platform with AI on the side of capturing the communications data, which you must do. You must capture all the data. You must capture it in a good and rich format. We've lit that up with AI. So now we don't just capture email, text, voice, video, all the things you would've known of, say, a year ago. Now we're also capturing Amazon Bedrock and Google Workspace, and of course, Claude for Enterprise, as well as any of the AI tools that our clients need because they need to be able to capture and monitor what's happening with their AI in order to keep it safe.

Gemma Allen

>> Well.

Kim Crawford Goodman

>> We've lit up our fundamental platform. We've made our platform fully comprehensive because many times our customers have pockets and silos of data and capability that speaks to their communications data and speaks to how they use it for compliance. We're pulling all of that into one very capable integrated platform and giving people more access to it. So we've lit up the fundamental platform. We're making that platform more useful with a whole suite of agents that I'm sure we'll talk about some because the AI is not just crawling the data now looking for bad behavior. Agents can work in the data and accomplish a lot on our behalf, on behalf of our customers. And last but certainly not least, we've lit up the platform through access. Of course, like every technology, we've had access for many years now with APIs. But since we were here in April, We've announced our MCP server, which allows people's agents to access actually our platform, the data and the agents on our platform. So there's so much that's happened in just 6 months.

Gemma Allen

>> Well, you're certainly meeting the market moment, that's for sure. There's so much to unpack there. I really want to talk about MCP and I want to talk about Agentic. But first, I want to start with something quite basic, right? And that is that in a short space of time, what it means to communicate and who's communicating has changed pretty dramatically. Last week we saw the rollout of Instinct. Are you familiar with this invite-only viral app that folks are downloading? You can use it through WhatsApp or text message and ask it to do all sorts of things for you. Book your holidays, talk to you, advise somebody at work you'll be late for a meeting. Right. And it's taken off so quickly, I almost couldn't believe it myself when I read the news piece this morning. So we know who's communicating with whom and when, it's become a whole new opportunity and conundrum. From your perspective, as all of these comms channels are exploding, what are your thoughts around how things are being governed at an enterprise level? What sorts of risks and compliance challenges have you guys seen?

Kim Crawford Goodman

>> Yeah, this is what you're articulating is one of the age-old challenges of communications and of compliance and of intelligence. That just continues to roar forward. What I mean by that is communications capabilities are always expanding. If we had been talking just a year or two ago, we would have been talking about the billions of dollars of fines that clients had received because all of the mobile apps had come to bear and you had investors communicating with their advisor or you had traders communicating with each other. On things like WhatsApp and WeChat and Signal, and it was not properly captured. And that resulted in a compliance failure and literally billions of dollars of fines. Now we see the explosion of AI capabilities. As I just mentioned at Smarsh, we are always first to market to capture what is actually going into this AI and what's coming out of it. And we know that there will be a need for that in the future. We talk to our clients all the time and they are always describing whether it's in Slack or something else that new technologies have just added a chat capability to that technology. Whenever these things are happening, it must be captured. It must be captured in rich and good format and it must be available for the intelligent capabilities of people like Smarsh to crawl. And now we're looking not just for risks and bad things, we're also looking for opportunity. At Smarsh, we describe that what we do more than anything is we preserve and protect the truth. And in this day and age, you must know this better than anyone as a reporter, what is actually true has— right, has— it's changed how you find what is truly truthful. And there's a lot in the atmosphere that is actually asserted but is not necessarily truthful. So it makes the capture, the use, the intelligent evaluation of this communications data particularly by regulated companies, by government industries, by financial companies, it makes it even more important and it's even more important for it to be completely comprehensive.

Gemma Allen

>> So if we think about the model a year or two ago, a year or two ago, within any enterprise, right, you need to capture data, you need to archive that data, you need to be able to search and discover that data, and then you need to surveil that data, right? Like that is absolutely essential to what needs to happen. In an agentic world, that needs to happen like a million X, really, right? Because we don't know how many instances of new data are going to be created and stored in any one environment on any given day, especially in a world where AI agents are actually accessing systems on behalf of humans. So when you guys go to build out this technology stack around that problem, where was the key focus first? Was it on, a better version of Capture? Was it on the surveillance end from the perspective of cyber? How has it kind of changed what you guys are building towards from a product perspective and where do you see this headed?

Kim Crawford Goodman

>> Yes, it's an excellent question. And what you described is the fundamental process for compliance. And it's also— you're also touching on why Smarsh and our scale is so advantageous to our clients. As in many industries, when you have larger scale, you can invest more, you can enrich more, and therefore you can add more value. So we are investing both in the compliance workflow that you just went through, but also how do you take the value of this communications data beyond compliance because it is so rich. On the compliance side that you just described of capturing, archiving, and surveilling, we are investing across all three. The capture has to get broader and broader. And we've introduced, as I mentioned, capture across all kinds of relevant AI platforms as we continue to chase every communications vehicle. We bought a company early last year that does voice, because what we see is that to really utilize this data properly in the future, peo— our clients are gonna want all communications together, not just the email and the text and the WhatsApp in one space, but they're going to want all of the voice and video all there together to be searchable. But we have invested in the surveillance because what happens in the business communications is that many times people are surveilling, they're looking for risk, but it's like looking for a needle in the haystack. And so it is very helpful with the right AI technologies to actually look for real risk It means that you are eliminating things called false positives. It means that you're making people's fundamental operation and compliance far more efficient because your false positives go down by 70, 80%. And what the resources are really being used for is to analyze and find real risk. And so we are investing so our clients can utilize their resources on real risk and not things that are false. But then we are investing outside of compliance as well and creating capabilities for this data, which is so rich, to be used in other areas. A great example of that is we have an agent, we have a whole suite of agents, but one of them is called Discovery Agent. It is for investigations. It's for legal work so that if there's some issue or dispute, say between Gemma and Kim, Instead of just doing a dump of every single piece of data from you and me, anything we've ever said, period, or to each other, now what can happen is the agent can go in there, the agent can crawl all kinds of communications, the agent can see the context. I've seen a demo of this and the agent can actually create a context graph of every time we talked to each other, but then every single person you talked to about me or I talked to about you, and you can see kind of a graph of how This investigation needs to proceed based on how communications expanded.

Gemma Allen

>> Wow.

Kim Crawford Goodman

>> So the investment we've made is tremendous on the compliance side to be at the very forefront, to work in a partnered way with our customers on our Smarsh Communications Intelligence Platform that is lit up by AI, but also to invest in a suite of agents that can help them use this technology and use this data to help their enterprise more broadly.

Gemma Allen

>> So I want to get on to MCP because I truly, when I looked at your business, I thought that I think that is a very smart move, right? You have to roll with the direction that this industry is headed. And we know that folks are going to design their own agents. They are going to want to call systems. So allowing them in, I think, is actually truthfully a great idea. Talk a little bit about what it means from your perspective for your TAM. You and I will have agents that are external coming in looking through. And your data, I'm sure, is like a huge part of the moat. It's a sticky nature of this business, right? It's a powerful piece that you bring. How did you think about that? When you guys are weighing up that decision to go forward with this MoE model, especially in the world of frontier models and LLMs. Talk me through what sorts of ways in which you guys debated that and where you landed on it.

Kim Crawford Goodman

>> Well, we're actually going to be talking to our customers directly about this tomorrow at our customer event that you mentioned. What we see very clearly, because we work across the industry in financial services as an example, we work with 18 of the 20 largest financial institutions outside of mainland China. What we see across the industry is that there are going to be all kinds of models used. So in the press, you all do talk quite a lot about the large language models, and they are very capable and they're a very important innovation. But what we see already happening with our clients and will happen more is they will— so they have access to and they will select from a variety of models. The large language models will be out there. But there are already open source models that depending on who you are and where you are, people are choosing from. But there are also a multitude of small language models. And we at Smarsh, because we've been involved in AI even before it was AI, they called it machine learning. So we've been involved with this for 10 years. We have actually refined a small language model all around communications data and what it can do on behalf of our clients. And because it is a small language model, it's very fit for purpose, both in terms of how good it is, but also how efficient it is from a cost standpoint. So what we see at Smarsh is people will develop and utilize all kinds of agents from different sources, and they will do that on the back of different types of models. So we are developing native agents to our platform that people can use. We are doing that on the back of our small language model. However, we see that our clients will develop their own agents, our clients will want to bring those agents to integrate with and utilize the quality data that we have because we haven't talked about it. But one of the things that's true in the world is not all data is the same. Much communications data is messy. Much communications data doesn't properly ascribe Who actually sent what and when did they do it? Our data is foolproof and pristine because it has to be used for legal cases, for regulatory, et cetera. So our clients will want to develop some agents. They'll want to come and use our small language model. They'll want to interact with our data. Their agents will eventually talk to our agents, and that has to be done in a sophisticated way through MCP servers. So that's why we were very keen to invest in this, very keen to develop it and bring— and be the first to bring it to market.

Gemma Allen

>> Like I said, I think it's a great idea because you need to meet customers where they're at. And this is a confusing moment in technology. A lot of folks are making buying decisions, not really knowing what 5 years out looks like. So understanding how you can continue to be part of their journey is huge. I want to ask you about the competitive landscape that we're in. we ask all folks about this.

Kim Crawford Goodman

>> Yes.

Gemma Allen

>> When I think about the surveillance element of your business as a good example, we've talked to a lot of folks in cyber. A lot of money being spent in R&D in that space. And sometimes I think about it from the perspective of convergence, right? We often debate with CEOs and founders on this show. Will we see a mass convergence? Will we see a world where, to quote an unnamed Frontier Lab leader, one orchestrator will do everything? How do you guys think about the competitive dynamics that you're in? You have companies like Microsoft and I know there's big players, smaller players who I'm sure are also coming for this long-tail opportunity. Right. I'd love to understand from you, Kim, how do you think about that and what does the next 2 to 5 year priority look like from that perspective?

Kim Crawford Goodman

>> You know, we have a range of competitors that we have a great deal of respect for that are formidable. I used to work for Michael Dell. And one of his philosophies at that time was it's good to have good competition. Yeah, because it keeps you on your toes and it makes you excellent. So, yes, we have competitors that are traditional and focused in our space, although a fraction of our size. We have competitors that are kind of new, funded, brand new and somewhat AI-oriented that are even smaller in size. But that, as you mentioned with the new applications, that can change very quickly. But what they don't have is the depth of capability of all the security that's needed, all of the interfaces that are needed. There's a whole— you almost can think about it as an iceberg. And there are very beautiful things at the top, demos that people can do, but there's so much capability that you have to have underneath to have the data right, to have the scalability, to have the global capability, to have the repeatability. And then we do have, large companies. Obviously Microsoft is a major communication provider. What we have found is the needs in this space for regulated companies go so deep that many times when people are either brand new or very broad, they don't satisfy the needs as deeply as we do. You asked me about our investment areas. We don't have 50, 100, 1,000 investment areas. We are always making the usefulness of this communications data better, and that is our absolute focus. And that's what keeps us ahead of competition.

Gemma Allen

>> Tell me about new opportunities you're seeing on the market. I know you guys have a huge lion's share of the financial services industry. Where else are you seeing new needs arising because of this agentic world we're all in?

Kim Crawford Goodman

>> Yeah, I would say first and foremost, we're seeing an expansion of people needing this quality of communications data. For example, in places like government. Government has a Freedom of Information Act. If they close the elementary school down the street from you, you have a right to go to that government entity and ask for every email, every, hearing, every document that existed that justified the closing of the elementary school where you thought you were going to send your son or daughter. Health care— we know that health care has all kinds of laws of appropriate privacy and appropriate decision-making that affects the life and death of all of us as individuals. So we see an expansion of the market from a TAM standpoint for the quality of this information. We also see an expansion of the market for the use of the information. So I described, for example, the Discovery Agent, which is a legal use that is not restricted to people who are regulated. We have a whole suite of agents that are— some are introduced and others that are coming on this communications data. A perfect example is we have an agent called AskSmarsh. This allows employees, just like you can query the internet in natural language methods, you can now query your own communications data in very natural language methods. And because this is private and pristine information, it's not accessible by the large language models. The way it's accessible is by the AskSmarsh agent. We see use cases for driving revenue growth over time from a CRM standpoint, because sitting in this information is everything that a customer asked for, how employees responded. And you can combine that with other datasets to see what actually enabled things to be sold versus not. So we see opportunity on the communications compliance side from new industries that have requirements like government and healthcare, pharma and others. But we also see tremendous opportunity across multiple industries for other use cases well beyond regulation and compliance.

Gemma Allen

>> Well, I think if you can offer enterprises and all customers access to structured data that's protected without the risks of data retention and all of those things, you certainly have a model that's needed in this moment. We're in right now. So, Kim, last question for you. I know you're here in New York. You're meeting some customers tomorrow at the Rockefeller Center. I guess what's your, what's your main message for this week? What do you want everyone to leave that room knowing tomorrow on behalf of you and the Smarsh exec?

Kim Crawford Goodman

>> My main message for the week is that your communications data is tremendously important and Smarsh is making it more and more intelligent.

Gemma Allen

>> I love that.

Kim Crawford Goodman

>> For you to use in ways that are more efficient and effective. If you already use it for compliance regulation, But it is also for use of things that are greatly expansive, whether that be legal, whether that be cyber-related, whether that be searching, whether that be gaining more revenue. So the opportunity— we are in a moment where the risks of being left behind are huge. Every entity, every professional has a risk of being left behind, but our clients don't need to worry about that because when they partner with us, we are deeply invested in keeping them at the forefront and giving them real value.

Gemma Allen

>> Well, I love it. It is certainly a moment we need to think as intelligently and as effectively as we can. So, Kim, thank you so much for joining us on NYSE Wired.

Kim Crawford Goodman

>> Thank you very much for having me.

Gemma Allen

>> I'm Gemma Allen here at theCUBE Studio at the New York Stock Exchange. This is NYSE Wired: Mixture of Experts. Thanks for watching.