theCUBE + NYSE Wired: Mixture of Experts Series | Anshul Gupta, Actively AI

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

Anshul Gupta

Co-Founder

Actively

play_circle_outline AI Agents Transform Revenue Teams: Free Sellers from 60–70% of Non‑Revenue Work

play_circle_outline Active learning: agents learn from collective salesperson actions and outcomes.

play_circle_outline Actively AI Raises $45M Series B, $250M Valuation to Offer Palantir-Style Forward-Deployed Enterprise Engineering

play_circle_outline Agent Memory as the New System of Record: Funding Drives Product Roadmap, Market Education and Go-to-Market Expansion

Info

Anshul Gupta, Actively AI

Anshul Gupta

Co-Founder Actively

search

Gemma Allen

>> I'm Gemma Allen with theCUBE here at our studio at the New York Stock Exchange. This is Mixture of Experts, one of our programs with NYSE Wired, and today we're talking about what a sales expert will look like 10 years from now. Joining me now is Anshul Gupta, president and co-founder of Actively AI. Welcome, Anshul.

Anshul Gupta

>> Thank you for having me back. Good to be here.

Gemma Allen

>> So you guys had some exciting news last week. You've just raised a Series B. Interesting company at a very interesting time. Really super intelligence for sales. Give me the lowdown. Set out the stall here. Tell me a little about this company and the journey you've been on.

Anshul Gupta

>> Absolutely. Yeah. And thanks again for having me. So maybe I'll zoom out and talk about why we're solving the problem that we're solving. If you sort of look at how AI is transforming industry by industry, we've seen some functions that have seen massive, massive productivity lifts and have been transformed for good. Coding is a great example. If you look at the output of programmers and software engineers through technologies like Cursor, Claude Code, Codex, you've seen huge efficiency gains. And because building product is one of the two unbounded problems and unlimited upside problems that companies can work on, that grow the pie, you've seen so much money, impact transformation go in there. The other unbounded problem in our view, an unlimited upside venture in our view is go to market, right? Which is selling your products to more and more customers to provide more and more and more value. And companies, as you know, are spending 30 to 40% of every dollar that they earn on sales and marketing. But if you look at the reality, even with all of the progress we've seen with AI and LLMs, sellers are still spending 60, 70% of their time on what we like to call non-revenue-generating activity. It's all the things before actually meeting or after actually meeting with the prospects and customers from research, from preparation, from figuring out the right things to communicate, from identifying risks, from forecasting, even updating your CRM. And so all of this time is spent. And the reality is that we struggle as salespeople with a human capacity problem in that there's so many things I could be doing in a given day, but I only have so much time. And so if I'm a sales rep with a territory of 200 accounts or even a hundred accounts, I'm only focusing on a couple at a time. There's probably even gaps in those companies that I'm focusing on and then the rest I'm not even thinking about. So what Actively comes in is we're really trying to break through that human capacity limit and transform revenue organizations with AI agents that are constantly working and thinking about every single account, maintaining all of that context and helping progress the deal forward. And so that's, and we call that intelligence-led revenue because it's a shift from this sort of human-constrained execution world to one where you now have more agents than sellers that are empowering the sellers to drive more productivity across pipeline generation, closing deals and expanding customers.

Gemma Allen

>> Wow. Okay. The name, Actively AI.

Anshul Gupta

>> Yes.

Gemma Allen

>> Talk to me about where that came from.

Anshul Gupta

>> Absolutely. So Actively is short for a concept called active learning, which is what my co-founder here worked on as thesis research on Stanford and very much informs how we think about AI agents and go to market. And the whole concept around active learning is how can you build a system that actually learns from experts and continuously and actively learns and improves over time. And so the idea is if you have a large sales team, there's so many actions that are being done every single day and so many outcomes that are being captured. How can, rather than each individual operating in a silo, we take those collective learnings to make everybody better? And we found that go to market was a really, really ripe domain to be able to do that because of how much unstructured data that's being captured in different systems, how many outcomes, whether micro level in terms of booking a meeting or closing a deal that are being captured. And so if you can combine all of that with proactive agents, we think that the sky's the limit for productivity you can drive.

Gemma Allen

>> When I think of some of the most successful salespeople I've worked with, I worked at Microsoft in d a sales team and I worked with some real sharks, right? They were like charismatic sharks. Some of the things that they were bad at was like admin, right? They were terrible at updating pipeline and forecast. That said, they never missed it though. Okay? They always delivered. They were incredibly good at charming people, and often that happened in person. And these are enterprise accounts, which is a specific sub-tier of sales. Obviously SDR, it's a more kind of all to many approach. But when we think about that and we think about the future of sales, how do you imagine a ratio of people who are super charming, charismatic, went to have alumni networks, all of that stuff that has really driven the world of sales as we've known it versus this agentic army of agents that are swarming, creating all of this top of funnel data, intelligence signals, connecting points, and drilling that down into something that's tactical for another agent potentially or a sales person to use. Where do those roads meet and where do they like, I guess, collide?

Anshul Gupta

>> Yeah. It's an excellent question. So Anthropic, I believe it was last month, maybe two months ago, published this kind of economic sort of labor market impact report. And what they did is they went through each of the different business functions within companies and looked at today, what percent of that role is the work done by AI or done by agents? And in like an AGI, futuristic world, what percent can it theoretically go to? If you look at engineering as an example, the theoretical limit is like 98, 97%, which means the vast, vast, vast majority of code being shipped and products being built are going to be driven by agents. Sales is actually really interesting because today it's around 15 to 20%, but the theoretical limit is only, in air quotes, 60%, but 60% is a huge transition from today. And so, what are the impacts of that based off the exact question you're asking? One is we are going to shift from it primarily being human-driven to agents helping with a lot of the work in the background, but that other 40%, it's the success there is going to boil down to exactly what you talked about, which is people that are really, really good at the human to human connection part of it. And all the other work, research, strategy, prioritization, multi-threading is going to be handled by agents. And especially down market, the agents are probably going to be able to do more and more, right? If it's, to your point, maybe the agents can sell to agents and so we can be the context layer that helps power that or maybe it can be higher velocity. But in very much in the enterprise context, to your point, I think those charismatic sellers that you worked with in this future world are going to be even more in demand and they're actually going to be able to produce even more revenue than they were before because now that skillset of them is multiplied, right? And that's what we think about like the whole notion of a per account agent that we're building is, how do I take the expertise of those bestsellers and farm that out at scale across my entire total addressable market?

Gemma Allen

>> So you've just raised 45 million Series B. Company is valued at 250 million now. So you guys, people see the value in this, right? So this interesting to, for the folks that are interested in being on your cap table, when you first set out to win the first set of customers, profile them for me. Are they mid-market customers? Are they folks that already have or can't hire SDRs fast enough or can't afford SDRs? Or are they, to my point earlier, these enterprise customers that want to take their bestsellers and create replicas of them?

Anshul Gupta

>> Absolutely.

Gemma Allen

>> Who is the ultimate buyer right now for you?

Anshul Gupta

>> Yeah. So we're laser-focused on scaling companies and enterprise companies because ultimately we're in the business of giving you the platform and foundation to increase revenue per rep. And so obviously the more sellers you have, the more complex your motions are. Many of these larger enterprises, as you know, multiple products, multiple geographies, multiple segments. Our agents are well-equipped to kind of handle that within enterprise environments where the data's messy, it's all over the place and they're able to kind of make sense of that. And we have Palantir-style Forward Deployed Engineering model where we work very, very closely in driving the customization of the account agents, the deployment, the change management of getting this in the hands of the sellers. And so we're really much equipped for the enterprises who really are the ones that have faced the problem most acutely because the kind of more old school Salesforce CRM model is kind of tapping out.

Gemma Allen

>> Talk to me about the tech. Are you guys building everything proprietary? Is this multimodal? I'm sure that all of this intelligence, all of these separate like graph signals that come from all of these connective points across enterprise, especially when you think about tacit knowledge, right? That comes from multiple tools and products right now.

Anshul Gupta

>> Absolutely.

Gemma Allen

>> What is the plan here? Are you guys building something that kind of consolidates everything? How do you think about that?

Anshul Gupta

>> Yes. Yeah. So in terms of the large language models, we work with kind of the standard providers for different parts of the underlying process. But yeah, in terms of the proprietary technology that we've built that's very difficult for internal teams, et cetera, to replicate and so they'd rather just build on top of our infrastructure is this notion of the per account agent, right? Where we have one context-dense agent that literally runs forever and lives with that account throughout the life cycle from prospect to customer to renewing to expanding. And to your point on the organizational level, right, the per account agents are connected to this broader kind of context graph that to your point is able to capture both the tacit knowledge and then also the continuous learning of when it recommends these actions to the sellers and they do them, here are the actions that are happening and then feeding that back in, going back to the name Actively to make this thing smarter and smarter and more productive and dynamic over time. And so those are sort of the areas that we're really, really investing in because that is the bar that you have to be able to achieve to be able to actually guide sellers on what to do differently.

Gemma Allen

>> I'm ramp. I'm like, "Yeah, this sounds great. I want to buy this product. I want to plug this in." Talk to me through what that deployment looks like and what the pricing looks like too. Is this consumption-based? Are you also part of that consumption way, right? We're told everything's going to be usage for token. How are you guys thinking about that side of the business and tokenomics generally, building in a world that's financially quite fragile and quite unpredictable?

Anshul Gupta

>> Absolutely. And so to get started, it's actually rather, rather simple. We have incredibly fast time to value where if you decide to work with us, we can scope a kind of pilot or land zone, turn on some number of account agents and you'll see value within a couple of weeks because we're able to integrate into all of the different systems wherever you decide to start with and can integrate more over time. And we're able to get it quickly into the hands of the reps and the sellers and deliver them actions that they should be taking that they're not, or accelerate the work that they have to be doing. And so the time to value is really, really quick and that's something that we pride ourselves on. In terms of the pricing model, I think that's also a great question. We are pretty increasingly sort of consumption-pilled. I think like per seat pricing largely is going to be dead, especially if technology like this works, you might not in some specific functions and departments need as many seats, right? And so I think that's one way that we're also countering against a lot of the more incumbent generic per seat pricing models. And so we kind of are able to price on per account, the number of account agents, the complexity and the scope of the account agents that are running in the background every day, as well as kind of usage parameters as well. And so you can grow into it, which is the nice part.

Gemma Allen

>> Talk to me about situations whereby companies have been relatively successful, but because they've had like a couple of sellers that have a lot of good relationships and a lot of knowledge in their head, right? Like I said, some of the best salespeople are not good at documenting anything, right? They're like hardcore hustlers. They don't want to be doing admin or even using Salesforce. That's a known problem. When you go out to like scope these companies and there's a discoverability, how do you guys plan for making sure that the data that's feeding the contextual layer of this plugin, of this opportunity is as strong and as effective and as relevant as it can be? Is there an element of trying to map relationships, understand historical sales context? How does it work in that side?

Anshul Gupta

>> Absolutely. There's a number of things that we do. The first is exactly what you said, which is you have to learn historically sort of what's working and develop a perspective on that. The second is to be context-dense and complete to provide value to the sellers, you have to be able to plug into all the different areas that there's context. I think the third and most interesting thing to your point is like if I were to go ask a sales leader at any company today, "Hey, Mr. And Ms. Sales Leader, what's the status of this deal?" The last place they would go today is Salesforce. Why? Because Salesforce's model, the CRM's model, to your point, it requires you entering information in and then it doesn't do anything for you. I have no incentive to put information into something that doesn't do anything. We actually change the incentive model, which is even sellers that use Actively, if they have an in person conversation with a customer or prospect that isn't recorded, maybe they'll take notes on their pen and paper. They actually upload that into the agent's memory. Why? Because the agent now does stuff for them. If there's follow-ups that come out of that meeting, the agent is going to go help you do those. If the relationship mapping gets updated because I learned, oh, hey, there's a new person getting hired and they're actually going to have more influence, the agent is actually going to update the actions that I have to do regardless. And so when you change the incentive model, you'd be surprised by how that drives capability and then it compounds the context that you're generating on those accounts as well, which is a benefit for the business too, because even if the seller leaves, you're maintaining that context for that account for forever. And so by changing the incentive model, you drive pretty remarkable outcomes.

Gemma Allen

>> So when I think about products like this and I imagine my career and when I would have used this or benefited from this, I like everything about it. I like the top of funnel, the intelligible signals, the connection, the iteration, like that's great, right? That data is powerful. The part I don't like is when a bot spams me on LinkedIn.

Anshul Gupta

>> Absolutely.

Gemma Allen

>> How do we manage that? Because people feel that, right? You know this is a bot and LinkedIn has become somewhat of a spam engine-

Anshul Gupta

>> 100%....

Gemma Allen

>> in my opinion. How do you get to the conversion? How do you make sure that that happens in a way that still feels valuable and has that even human connection point, which I think a lot of people still enjoy is in 2026. 2036, maybe not, but for now, when agents are selling to agents, it's another challenge, right? But at this point, how do you think about that?

Anshul Gupta

>> 100%. I mean, one thing is we've trained our kind of content writing capabilities on a ton of historical sort of information that helps it be more attuned and continuously updating. The other beautiful part that we have within our agents is they're able to have both user and company-level memory. The user-level memory is exactly what you said. You might have a different way of communicating with people than I do, right? I'm very short, straight to the point. I have brackets around my numbers. I have a very, very particular style of communicating and my account agents for my territory are actually able to learn that, but then infuse all of the relevant context that's needed that I'm going to communicate to the person about, "Hey, this is why you should take a look at the problem that I'm solving or this is what I've learned talking to other people about your business and why it matters to you." So if you can pair all of the rich context, the right POV and the right timing with that human element of your particular communication style, that's what I think the most powerful combination is. And then we're seeing that with our customers today.

Gemma Allen

>> Okay. Last question.

Anshul Gupta

>> Yeah.

Gemma Allen

>> 45 million in the bank.

Anshul Gupta

>> Yes.

Gemma Allen

>> What's ahead? What are you spending that money on? Where are you doubling down?

Anshul Gupta

>> It's a really, really good question. I think for us, it's really transforming the industry. What I believe is 100% inevitable and going to happen is that within the next two or three years, every go-to-market organization is going to be powered by these agents that are telling them, their teams, "This is what you should be focused on," and helping them do them, and that's going to result in this productivity gain. So I know that that's going to happen. And so the question is, with our funding, how can we educate the market on the change that's going to come? How can we deliver and continue to compound on our ambitious product, product roadmap where we think you don't actually need all of these tools in the future. You might not actually need a CRM in five years because we think the agent memory is going to become the system of record. And so we have an ambitious product roadmap, kind of ambitious roadmap in educating the market and then continuing to grow our own go to-market efforts. We call it Actively Un-Actively, right? Our teams religiously kind of use and help our product become better as well.

Gemma Allen

>> And afford the rent here in New York City, right?

Anshul Gupta

>> Exactly, exactly.

Gemma Allen

>> Anshul, thank you so much for coming on theCUBE.

Anshul Gupta

>> Thank you for having me.

Gemma Allen

>> I'm Gemma Allen here at theCUBE Studio at the NYSE. This is Mixture of Experts, our program with NYSE Wired. Thanks for watching.

Anshul Gupta, Actively AI

search

Gemma Allen

>> I'm Gemma Allen with theCUBE here at our studio at the New York Stock Exchange. This is Mixture of Experts, one of our programs with NYSE Wired, and today we're talking about what a sales expert will look like 10 years from now. Joining me now is Anshul Gupta, president and co-founder of Actively AI. Welcome, Anshul.

Anshul Gupta

>> Thank you for having me back. Good to be here.

Gemma Allen

>> So you guys had some exciting news last week. You've just raised a Series B. Interesting company at a very interesting time. Really super intelligence for sales. Give me the lowdown. Set out the stall here. Tell me a little about this company and the journey you've been on.

Anshul Gupta

>> Absolutely. Yeah. And thanks again for having me. So maybe I'll zoom out and talk about why we're solving the problem that we're solving. If you sort of look at how AI is transforming industry by industry, we've seen some functions that have seen massive, massive productivity lifts and have been transformed for good. Coding is a great example. If you look at the output of programmers and software engineers through technologies like Cursor, Claude Code, Codex, you've seen huge efficiency gains. And because building product is one of the two unbounded problems and unlimited upside problems that companies can work on, that grow the pie, you've seen so much money, impact transformation go in there. The other unbounded problem in our view, an unlimited upside venture in our view is go to market, right? Which is selling your products to more and more customers to provide more and more and more value. And companies, as you know, are spending 30 to 40% of every dollar that they earn on sales and marketing. But if you look at the reality, even with all of the progress we've seen with AI and LLMs, sellers are still spending 60, 70% of their time on what we like to call non-revenue-generating activity. It's all the things before actually meeting or after actually meeting with the prospects and customers from research, from preparation, from figuring out the right things to communicate, from identifying risks, from forecasting, even updating your CRM. And so all of this time is spent. And the reality is that we struggle as salespeople with a human capacity problem in that there's so many things I could be doing in a given day, but I only have so much time. And so if I'm a sales rep with a territory of 200 accounts or even a hundred accounts, I'm only focusing on a couple at a time. There's probably even gaps in those companies that I'm focusing on and then the rest I'm not even thinking about. So what Actively comes in is we're really trying to break through that human capacity limit and transform revenue organizations with AI agents that are constantly working and thinking about every single account, maintaining all of that context and helping progress the deal forward. And so that's, and we call that intelligence-led revenue because it's a shift from this sort of human-constrained execution world to one where you now have more agents than sellers that are empowering the sellers to drive more productivity across pipeline generation, closing deals and expanding customers.

Gemma Allen

>> Wow. Okay. The name, Actively AI.

Anshul Gupta

>> Yes.

Gemma Allen

>> Talk to me about where that came from.

Anshul Gupta

>> Absolutely. So Actively is short for a concept called active learning, which is what my co-founder here worked on as thesis research on Stanford and very much informs how we think about AI agents and go to market. And the whole concept around active learning is how can you build a system that actually learns from experts and continuously and actively learns and improves over time. And so the idea is if you have a large sales team, there's so many actions that are being done every single day and so many outcomes that are being captured. How can, rather than each individual operating in a silo, we take those collective learnings to make everybody better? And we found that go to market was a really, really ripe domain to be able to do that because of how much unstructured data that's being captured in different systems, how many outcomes, whether micro level in terms of booking a meeting or closing a deal that are being captured. And so if you can combine all of that with proactive agents, we think that the sky's the limit for productivity you can drive.

Gemma Allen

>> When I think of some of the most successful salespeople I've worked with, I worked at Microsoft in d a sales team and I worked with some real sharks, right? They were like charismatic sharks. Some of the things that they were bad at was like admin, right? They were terrible at updating pipeline and forecast. That said, they never missed it though. Okay? They always delivered. They were incredibly good at charming people, and often that happened in person. And these are enterprise accounts, which is a specific sub-tier of sales. Obviously SDR, it's a more kind of all to many approach. But when we think about that and we think about the future of sales, how do you imagine a ratio of people who are super charming, charismatic, went to have alumni networks, all of that stuff that has really driven the world of sales as we've known it versus this agentic army of agents that are swarming, creating all of this top of funnel data, intelligence signals, connecting points, and drilling that down into something that's tactical for another agent potentially or a sales person to use. Where do those roads meet and where do they like, I guess, collide?

Anshul Gupta

>> Yeah. It's an excellent question. So Anthropic, I believe it was last month, maybe two months ago, published this kind of economic sort of labor market impact report. And what they did is they went through each of the different business functions within companies and looked at today, what percent of that role is the work done by AI or done by agents? And in like an AGI, futuristic world, what percent can it theoretically go to? If you look at engineering as an example, the theoretical limit is like 98, 97%, which means the vast, vast, vast majority of code being shipped and products being built are going to be driven by agents. Sales is actually really interesting because today it's around 15 to 20%, but the theoretical limit is only, in air quotes, 60%, but 60% is a huge transition from today. And so, what are the impacts of that based off the exact question you're asking? One is we are going to shift from it primarily being human-driven to agents helping with a lot of the work in the background, but that other 40%, it's the success there is going to boil down to exactly what you talked about, which is people that are really, really good at the human to human connection part of it. And all the other work, research, strategy, prioritization, multi-threading is going to be handled by agents. And especially down market, the agents are probably going to be able to do more and more, right? If it's, to your point, maybe the agents can sell to agents and so we can be the context layer that helps power that or maybe it can be higher velocity. But in very much in the enterprise context, to your point, I think those charismatic sellers that you worked with in this future world are going to be even more in demand and they're actually going to be able to produce even more revenue than they were before because now that skillset of them is multiplied, right? And that's what we think about like the whole notion of a per account agent that we're building is, how do I take the expertise of those bestsellers and farm that out at scale across my entire total addressable market?

Gemma Allen

>> So you've just raised 45 million Series B. Company is valued at 250 million now. So you guys, people see the value in this, right? So this interesting to, for the folks that are interested in being on your cap table, when you first set out to win the first set of customers, profile them for me. Are they mid-market customers? Are they folks that already have or can't hire SDRs fast enough or can't afford SDRs? Or are they, to my point earlier, these enterprise customers that want to take their bestsellers and create replicas of them?

Anshul Gupta

>> Absolutely.

Gemma Allen

>> Who is the ultimate buyer right now for you?

Anshul Gupta

>> Yeah. So we're laser-focused on scaling companies and enterprise companies because ultimately we're in the business of giving you the platform and foundation to increase revenue per rep. And so obviously the more sellers you have, the more complex your motions are. Many of these larger enterprises, as you know, multiple products, multiple geographies, multiple segments. Our agents are well-equipped to kind of handle that within enterprise environments where the data's messy, it's all over the place and they're able to kind of make sense of that. And we have Palantir-style Forward Deployed Engineering model where we work very, very closely in driving the customization of the account agents, the deployment, the change management of getting this in the hands of the sellers. And so we're really much equipped for the enterprises who really are the ones that have faced the problem most acutely because the kind of more old school Salesforce CRM model is kind of tapping out.

Gemma Allen

>> Talk to me about the tech. Are you guys building everything proprietary? Is this multimodal? I'm sure that all of this intelligence, all of these separate like graph signals that come from all of these connective points across enterprise, especially when you think about tacit knowledge, right? That comes from multiple tools and products right now.

Anshul Gupta

>> Absolutely.

Gemma Allen

>> What is the plan here? Are you guys building something that kind of consolidates everything? How do you think about that?

Anshul Gupta

>> Yes. Yeah. So in terms of the large language models, we work with kind of the standard providers for different parts of the underlying process. But yeah, in terms of the proprietary technology that we've built that's very difficult for internal teams, et cetera, to replicate and so they'd rather just build on top of our infrastructure is this notion of the per account agent, right? Where we have one context-dense agent that literally runs forever and lives with that account throughout the life cycle from prospect to customer to renewing to expanding. And to your point on the organizational level, right, the per account agents are connected to this broader kind of context graph that to your point is able to capture both the tacit knowledge and then also the continuous learning of when it recommends these actions to the sellers and they do them, here are the actions that are happening and then feeding that back in, going back to the name Actively to make this thing smarter and smarter and more productive and dynamic over time. And so those are sort of the areas that we're really, really investing in because that is the bar that you have to be able to achieve to be able to actually guide sellers on what to do differently.

Gemma Allen

>> I'm ramp. I'm like, "Yeah, this sounds great. I want to buy this product. I want to plug this in." Talk to me through what that deployment looks like and what the pricing looks like too. Is this consumption-based? Are you also part of that consumption way, right? We're told everything's going to be usage for token. How are you guys thinking about that side of the business and tokenomics generally, building in a world that's financially quite fragile and quite unpredictable?

Anshul Gupta

>> Absolutely. And so to get started, it's actually rather, rather simple. We have incredibly fast time to value where if you decide to work with us, we can scope a kind of pilot or land zone, turn on some number of account agents and you'll see value within a couple of weeks because we're able to integrate into all of the different systems wherever you decide to start with and can integrate more over time. And we're able to get it quickly into the hands of the reps and the sellers and deliver them actions that they should be taking that they're not, or accelerate the work that they have to be doing. And so the time to value is really, really quick and that's something that we pride ourselves on. In terms of the pricing model, I think that's also a great question. We are pretty increasingly sort of consumption-pilled. I think like per seat pricing largely is going to be dead, especially if technology like this works, you might not in some specific functions and departments need as many seats, right? And so I think that's one way that we're also countering against a lot of the more incumbent generic per seat pricing models. And so we kind of are able to price on per account, the number of account agents, the complexity and the scope of the account agents that are running in the background every day, as well as kind of usage parameters as well. And so you can grow into it, which is the nice part.

Gemma Allen

>> Talk to me about situations whereby companies have been relatively successful, but because they've had like a couple of sellers that have a lot of good relationships and a lot of knowledge in their head, right? Like I said, some of the best salespeople are not good at documenting anything, right? They're like hardcore hustlers. They don't want to be doing admin or even using Salesforce. That's a known problem. When you go out to like scope these companies and there's a discoverability, how do you guys plan for making sure that the data that's feeding the contextual layer of this plugin, of this opportunity is as strong and as effective and as relevant as it can be? Is there an element of trying to map relationships, understand historical sales context? How does it work in that side?

Anshul Gupta

>> Absolutely. There's a number of things that we do. The first is exactly what you said, which is you have to learn historically sort of what's working and develop a perspective on that. The second is to be context-dense and complete to provide value to the sellers, you have to be able to plug into all the different areas that there's context. I think the third and most interesting thing to your point is like if I were to go ask a sales leader at any company today, "Hey, Mr. And Ms. Sales Leader, what's the status of this deal?" The last place they would go today is Salesforce. Why? Because Salesforce's model, the CRM's model, to your point, it requires you entering information in and then it doesn't do anything for you. I have no incentive to put information into something that doesn't do anything. We actually change the incentive model, which is even sellers that use Actively, if they have an in person conversation with a customer or prospect that isn't recorded, maybe they'll take notes on their pen and paper. They actually upload that into the agent's memory. Why? Because the agent now does stuff for them. If there's follow-ups that come out of that meeting, the agent is going to go help you do those. If the relationship mapping gets updated because I learned, oh, hey, there's a new person getting hired and they're actually going to have more influence, the agent is actually going to update the actions that I have to do regardless. And so when you change the incentive model, you'd be surprised by how that drives capability and then it compounds the context that you're generating on those accounts as well, which is a benefit for the business too, because even if the seller leaves, you're maintaining that context for that account for forever. And so by changing the incentive model, you drive pretty remarkable outcomes.

Gemma Allen

>> So when I think about products like this and I imagine my career and when I would have used this or benefited from this, I like everything about it. I like the top of funnel, the intelligible signals, the connection, the iteration, like that's great, right? That data is powerful. The part I don't like is when a bot spams me on LinkedIn.

Anshul Gupta

>> Absolutely.

Gemma Allen

>> How do we manage that? Because people feel that, right? You know this is a bot and LinkedIn has become somewhat of a spam engine-

Anshul Gupta

>> 100%....

Gemma Allen

>> in my opinion. How do you get to the conversion? How do you make sure that that happens in a way that still feels valuable and has that even human connection point, which I think a lot of people still enjoy is in 2026. 2036, maybe not, but for now, when agents are selling to agents, it's another challenge, right? But at this point, how do you think about that?

Anshul Gupta

>> 100%. I mean, one thing is we've trained our kind of content writing capabilities on a ton of historical sort of information that helps it be more attuned and continuously updating. The other beautiful part that we have within our agents is they're able to have both user and company-level memory. The user-level memory is exactly what you said. You might have a different way of communicating with people than I do, right? I'm very short, straight to the point. I have brackets around my numbers. I have a very, very particular style of communicating and my account agents for my territory are actually able to learn that, but then infuse all of the relevant context that's needed that I'm going to communicate to the person about, "Hey, this is why you should take a look at the problem that I'm solving or this is what I've learned talking to other people about your business and why it matters to you." So if you can pair all of the rich context, the right POV and the right timing with that human element of your particular communication style, that's what I think the most powerful combination is. And then we're seeing that with our customers today.

Gemma Allen

>> Okay. Last question.

Anshul Gupta

>> Yeah.

Gemma Allen

>> 45 million in the bank.

Anshul Gupta

>> Yes.

Gemma Allen

>> What's ahead? What are you spending that money on? Where are you doubling down?

Anshul Gupta

>> It's a really, really good question. I think for us, it's really transforming the industry. What I believe is 100% inevitable and going to happen is that within the next two or three years, every go-to-market organization is going to be powered by these agents that are telling them, their teams, "This is what you should be focused on," and helping them do them, and that's going to result in this productivity gain. So I know that that's going to happen. And so the question is, with our funding, how can we educate the market on the change that's going to come? How can we deliver and continue to compound on our ambitious product, product roadmap where we think you don't actually need all of these tools in the future. You might not actually need a CRM in five years because we think the agent memory is going to become the system of record. And so we have an ambitious product roadmap, kind of ambitious roadmap in educating the market and then continuing to grow our own go to-market efforts. We call it Actively Un-Actively, right? Our teams religiously kind of use and help our product become better as well.

Gemma Allen

>> And afford the rent here in New York City, right?

Anshul Gupta

>> Exactly, exactly.

Gemma Allen

>> Anshul, thank you so much for coming on theCUBE.

Anshul Gupta

>> Thank you for having me.

Gemma Allen

>> I'm Gemma Allen here at theCUBE Studio at the NYSE. This is Mixture of Experts, our program with NYSE Wired. Thanks for watching.