Who Owns the Learning Curve?—Context Graphs and the New Enterprise Moat

Julia Nimchinski:

Next up… We’re transitioning to our next fireside Chat. Please welcome Jaya Gupta, partner at Foundation Capital. She’s nearly single-handedly responsible for the network effects around context graphs and context graphs becoming category. Her thesis reached 5 million views and was amplified by Satya Nadella. We’re so excited to feature you, Jaya, and Fireside Chat will be moderated by Seth Marrs, Chief Strategy Officer at Sandler. How are you doing?

Seth Marrs:

Dude.

Jaya Gupta:

Good, thanks for having us.

Julia Nimchinski:

Cool. Let’s get into it.

Seth Marrs:

Awesome. Hey, Jaya, how’s it going?

Jaya Gupta:

Good, how are you?

Seth Marrs:

Good, good, good. I’m super excited to be able to talk to you and go through this stuff together. It was interesting, I was on a CRO panel two days ago, and went through and asked what each person was doing, and every single one of the CROs said building and maintaining a context graph was their number one priority to have an advantage in AI. So… very cool to see the… I bet it’s really cool to see the impact of, kind of, some of the thought work that you’ve put into this.

Jaya Gupta:

Yeah, that’s awesome. I feel like it’s definitely, it’s been interesting. I think with how fast, sort of, like, news travels, and I know Jay just mentioned Twitter, it’s, it’s interesting to see, you know, when it actually comes into actual commercial practice, so I’m excited that people are talking about it.

Seth Marrs:

Yeah, cool. So I’m gonna just jump in and ask a couple questions around this, because it’s cool the way that you’ve put it together, the way that you’ve talked about it, these different types of data that have existed. So, just jumping in, like… Let’s start with your, like, the core thesis that you have. You separate, kind of, the rules, which I kind of interpret as the structured data, from things like official… like, it would be called something like using official ARR for reporting, from, like, decision traces, which are the pathway that you took that led to the decision that led to a result.

So… All the enterprise systems capture the rule, but none of them capture the trace. Why do you think that’s not happening?

Jaya Gupta:

Yeah, it’s a good question. I think system of records have always, stored, you know, kind of like, as you said, like, rules and, like, the, kind of, like, literal, definition of something. And you can… and this is, like, you think of all your classic, tools that, you know, everyone uses, whether it’s Salesforce. It’s, you know, everyone knows it’s built on, kind of, current state storage. And so you couldn’t get a sense of what the opportunity is. But I think the thing that is not being captured is, like. You know, all the things that sit in between different, like, organizations.

So, like, sales obviously communicates with finance, sales communicates with marketing. There’s, like, 18 different groups that all have different incentives, and I think a lot of the context is lost between these different groups. And a lot of these different groups are built around different system of records that have emerged for them over time. I think the thing that they’re missing, as you pointed out, is the idea of decision traces. And, you know, in very, very easy words, it kind of comes down to being a few things. It’s like. what are… what is the kind of exception logic that lives in, people’s heads?

It might be, hey, like, this company, we give, these types of customers a 10% extra discount, or 10%, kind of like… We even upcharge them. It could be precedent from past decisions. So, like, how do you think about, you know, we did… we structured a certain deal this way for one of our special customers, like, you know, yes, that’s not a rule, but how do we make that option? So, yeah, there’s a bunch of different use cases and things that tie into decision traces, which I think people have started to, like, look at now.

Seth Marrs:

Like, do you attribute that? Because all those come from conversations and communications that happen throughout a business that have… that are basically unstructured. And in a lot of cases, not even captured, right? Like, there is no me… you have it in an email, but… even companies today, there are very few that are using conversation intelligence, like, deeply to capture all of their conversations, internal, external. Like, is that the reason why this has never been… it’s just, like, nobody knew how to go do that, because the things you talked about are technically rules. But they’re rules that aren’t really applied, they’re kind of a whole bunch of these exceptions that have happened over… across a business decision when push comes to shove in a deal.

Jaya Gupta:

Yeah, yeah, yeah, yeah, it’s a good question. It’s a great question. I think that, you know, what… maybe… yeah, so you’re right that it is, like, you know, those are exceptions and the why part. I think, the part that I think is, like, really, really interesting, that, you know, you’re kind of alluding to is that people haven’t actually moved in towards conversational intelligence. And so, if you take all the Silicon Valley native companies. They use things like granola. They’re probably granola-ing you in real life. It’s like some recorder that doesn’t inform you that you’re being recorded.

They’re using things like WhisperFlow. Everything is essentially being documented, and I think that you’ve seen a lot of companies, like even RAMP and others, say that they’re moving towards a, like, everything’s sort of being recorded. Now, of course, that has a lot of different implications, and I think that the rest of the world probably is going to move in that direction, given the companies that can move in that direction will move in that direction. There’s companies that won’t be able to do it because of privacy and regulations and things like that, and they shouldn’t do it.

But yes, a lot of the conversation communication that’s lost is, like, is usually conversational, but I think that there’s also other different data sources as well that exist that… people, you know, weren’t tapping into. And LLMs can now parse unstructured data as well, so all your Notion documents, all your planning documents, every, you know, emails, you can connect it to almost everything.

Seth Marrs:

Yeah. How are you pulling… like, when you think… is there technology today that actually is structuring that in a way that it can be connected together? Because, I mean, it is unstructured, but the things you’re talking here are giving structure to that unstructured data and, like, pulling it together. So if you’re someone listening, going, okay, how do I build a tech stack around this? That is legitimately gonna help, because everybody’s saying they’re doing context graphs. My bet is when you look into what they’re doing, it’s not really what you’re saying a context graph is.

Jaya Gupta:

Totally. I completely agree. I mean, this is, like, the… you know, I work in venture, and the hard part is, like, diligencing companies, and thankfully, like, we talked about the concept and have a few companies that are… had actually been building it. So, I think that right now, the context graphs that are really, really good are ones where, you kind of have to go off of, like, doing a POC or working with them for 30 days, and you have to give them a good amount of data. And so I think that that’s, like, a big shift for people.

Like, this data also is not data that is going to come from just your team. It’s likely going to come from teams that are… are… don’t want to give the data, because they’re also worried about, like. You know, everyone has different incentives, as we talked about. And so I think that this is going to… context graphs are, one, an organization problem. Two, I think they’re also a tech… tech problem, and that, you know, there’s a company called Actively AI. I think they do a great job of building, like, a context graph underneath. And, you know, one of the reasons I, had insight into writing that article was actually, I would say, 3 companies.

Actively AI, which is building and sales, Max Moore, which is doing it for finance, and then Player Zero, which is building context graph for engineering. So, I think you have to find right now, we’re in an era where there is people that are combining, like. verticals that maybe, or functions that touch each other. And, like, I think that the trap here is that there’s context graphs that are probably saying, we’re gonna capture it for the entire company. I just don’t think that that’s realistically possible, starting from day one.

Seth Marrs:

Yeah, how could it be? Like, there’s so much information from coming from Sonar, and if I understand it right, like, it takes actually some intelligence to structure that, so it’s… it’s unstructured, but you’ve got to look through that using AI to be able to help you parse out, tune, and manage this in a way that it gets you to the result you want. And the way you talked about those three. They’re very focused on 3 different areas, and 3 different… Pieces of the business. That I’m assuming have to be tuned in 3 different ways.

Jaya Gupta:

Yes, exactly, exactly. And as you actually said, in finance, accuracy matters a lot. You can’t have, like. the idea of, like, hallucinations, and, like, I think finance people will take a much slower response from the model if it is accurate. And so there’s also, like, a trade-off of, like, how you are leveraging the model’s latency, and salespeople, they might want a really, really fast response. A lot of people, a lot of the things that they do are, like, urgent. And it’s like, if you can write me the best email in the world, like, in a few seconds.

With the right context, or even update me right in time as I’m jumping between meetings really, really fast. Speed is something that matters, and… for them. But there’s also a big difference. In sales, it’s really hard to figure out, like, why something happened. Like, why did a deal actually, like, go through? That part is really, really challenging. And then for finance, it’s much easier to figure out, like. you know, accounting. It’s verifiable. Like, you’re closing the books. All the evidence pretty much exists, especially in regulated industries for, like, accounting and finance, typically, because You even have things… you can now even have the models, like, reverse engineer Excel, like, formulas that finance people have put in.

Like, why do they put that formula in? You can have the model, even go through and audit every single model, reverse engineer a finance person’s logic. Which is what Max Moore does, to, like, reconstruct the context. So, it has to be verticalized, and it has to be different per, approach.

Seth Marrs:

And, like, the way you’re saying it, too, is it also isn’t about unstructured or structured data, it’s just about finding those things that are unique in individual situations. So, in a… highly ac… if you’re required to be highly accurate, you may apply a lot of this logic to very structured data, but just in using different tools in different ways. Where if you’re on the sales side, where it’s a lot of communication, you may be willing to have less resolution, because it’s giving you things that you could never see before.

Jaya Gupta:

Exactly, exactly. And I think that in sales, it’s really hard to judge accuracy. How do you know if an email’s perfect, or things… and sales will always be a moving target, which is the challenge, and also the blessing.

Seth Marrs:

Yeah, yeah, yeah, exactly.

Jaya Gupta:

Then we’re working.

Seth Marrs:

Yeah, when you talked a little bit about a few companies that are doing this well, and you mentioned Databricks having an advantage right now. Can you just go a little bit deeper into, kind of, why you think they’re a little bit ahead of the pack when it comes to doing this type of work?

Jaya Gupta:

Yeah, it’s a good question. I think, I think Databricks, one is just because of the way the company operates, and, I think that, they’re, they’re not, you know, they’re not, they’re not the type of company that wants to lose and will lose, and they’ve always sort of still maintained, somehow, their, like, AI research lineage. Even after being, you know, making so many, you know, billions and millions. And so I think they’ve been willing to innovate. I think that they’ve launched a bunch of new products as well, quickly, like Unity Catalog, and they’re working on, like, routing, too, and so I think that through all of this, they’re figuring out a way to make sure that they’re in, like, the right path of information.

And… and I think that, you know, you know, Foundation’s obviously investors in the company, so I have a biased take, but… So I have to throw that out there, but I just think that, they’ve done a remarkable job repositioning themselves as, like, you know, sort of agent-first and, you know, leveraging the data and sitting in the right path of agents.

Seth Marrs:

Yeah, I mean, the reason I ask, I get the bias, but you’re not the only one that has said that. I’ve heard that from multiple different sources around just that ingenuity and the way they’re kind of keeping themselves on the front end to be able to do things that other companies, or being willing to do things to stay on top of, like, almost running past the money now for money later. Like, trying to build to the bigger picture, it kind of feels like.

Jaya Gupta:

Exactly, exactly. And I think Databricks is positioning themselves as, hey, you’re going to build all your agents here. I think a lot of other companies are doing that, too. Glean is doing that, there’s tons of others, but I think that, you know, saying that you’re willing to do it, and then, you know, kind of, like, letting go of what you’ve thought of in the past, is the right way to innovate.

Seth Marrs:

Yeah, makes total sense. And not… and very difficult to do, too, right? Everyone talks about being able to do that, but actually doing that can be really painful.

Jaya Gupta:

Exactly, I completely agree. I mean, this is seen everywhere. I think everyone thought the incumbents would sort of… you know, I remember 3 or 4 years ago, everyone was, like, always worried about the incumbents, but there’s not a… even if it comes to context graph, like. like, you know, many of these incumbents, they reached out, being like, shoot, we need to build this, we have no idea how to build it, we’re gonna start selling it.

Seth Marrs:

Yeah.

Jaya Gupta:

We don’t know, we haven’t built it yet. And so, I think that this always happens where, you know, startups are kind of, like, at the cutting edge, and, I think here that the opportunity will, you know, hopefully be won by startups.

Seth Marrs:

Yeah, well… why wouldn’t… like, in the old world, which I think, obviously, AI disrupts pretty heavily. an SAP or a Salesforce would just buy the market, and they would be the first, because they could put the resources and energy into it to be able to make it real faster, but it doesn’t seem like that’s happening anymore. It seems like they… they’re almost… it almost feels like they’re disadvantaged, and maybe that goes back to the previous discussion we just had around Databricks. Yeah. Like, they’ve got a massive amount of money they’re generating. Like, what’s keeping them from dump?

They have more money than every… than all of these startups, they, like, why can’t they just take this and make it real?

Jaya Gupta:

Yeah, yeah, yeah, yeah. I think the problem is that context graphs are very, are very verticalized, and I think that they’re also very… and, you know, maybe this isn’t Databricks per se, but this is, like, probably true for, like, Microsoft and ServiceNow. Like, I think that context graphs, people don’t realize, like, I could talk about them technically all day, but I think the real challenge is actually, like, politically and organizationally. They’re very… because it’s like, everywhere, the spots where they would come to your head of where they’d be useful require getting buy-in from, like, 3 different groups.

And getting buy-in from those 3 different groups, Microsoft might have a good relationship with one of them. ServiceNow also probably has a great relationship with IT. Same with Microsoft, but… like, the people that are, you know, actually getting the, you know, most value out of these context graphs are typically the ones sitting in sales, or sitting in finance, or sitting in, like, the edges of the company versus, like, the central part. And so. data obviously touches, IT, but how do you get, you know, for DevOps, how do you get IT support and security ops aligned?

For, you know, something around, like, technical customer support, you want customer support, engineering, and product. And so, I think that the hardest Problem is, like, no one owns relationships with… VCs have always said, like, to startups, do not ever sell to 2-3 buyers at once.

Seth Marrs:

But… Yep.

Jaya Gupta:

A context graph breaks, like, literally every rule that most people have ever heard of in that domain.

Seth Marrs:

I mean, to a certain extent, that’s a very… it’s a hard and fast rule, because it’s the same thing. The minute, like, when I would talk to companies, and they start talking about selling into the sales and marketing persona at the same time, it’s like, good luck with that. It’s two different personas, and it’s… you can’t use one to get… it’s not as easy as it sounds. It’s all go-to-market, so people think that it’s easy.

In this situation, it seems like the organizational structure is the biggest challenge to doing this, because I would bet that sales, like in this example of sales and marketing, they have their own context graph project, they’re all running it with their own tech stack for doing it, and then marketing has their own that they’re doing, and neither of them will have any interest at all in bringing it together in one, even though it would be mutually beneficial, because it’s almost like a power play. And you’re talking about it… what you’re talking about is now you gotta go to the whole Oregon finance and…

Jaya Gupta:

Yeah.

Seth Marrs:

Can that problem be resolved, or is something gonna have to either break, change, or move? Because I don’t know how that… If all the technology’s there, how do you overcome that? Because those are pretty hard and fast walls that are built within organizations.

Jaya Gupta:

Yeah, it’s a good question. I think this is why, like, I think VCs have been, you know, worried about… not worried, but I think we haven’t seen, like, the Harvey for sales yet. And I think one of the reasons being is exactly what you articulated, in that, like, even just getting through sales and marketing is a problem. Now you have to talk about IT. getting involved, data getting involved, like, and hundreds of other parties, RevOps, CSMs. Because you need their context, too, to make it really valuable. I think the thing that will change it is probably, like, what you said actually in the beginning of this call gives me a lot of hope.

If you’re on a panel with a bunch of CROs, and now they’re starting to get it and say that this is their top priority, I think that what I would hope is that it actually is, a top priority, and that… I am starting to see you pick up, I will say, for the first 6 months of this article, for sure. I think Microsoft services… most of the incumbents hopped on. And I think that… What you saw from, like. sales is that they’re always a little bit more behind, and I think now it’s starting to have a massive pickup again.

Everyone’s realizing context is the answer.

Seth Marrs:

I mean, in sales, context is almost everything. The unstructured part is probably the least useful part of data in a sales organization, because it’s… not updated correctly. Now, when you bring in all of the unstructured stuff, the context is just… exposes everything. The sales rep who finds a way to get discounting nobody else does, the person who finds… like, all that stuff becomes very real in a world of context graphs.

Jaya Gupta:

For sure, for sure. And I think… I think the other thing that’s interesting is, like, a lot of salespeople are not even incentivized to… like, of course, there’s never been an incentive to put the data in, yes, it takes time, and all those things, but the other thing is, like, it’s just the incentive structure for sales is so different than any other role at most companies.

Seth Marrs:

That’s true.

Jaya Gupta:

And so, like, you know, if you’re a salesperson, you’re leaving your company, you’re quitting, going somewhere else, you want to take your contacts with you. And, like, you don’t… if it’s never in the CRM in the first place, your relationships, then that’s probably the best thing, because your relationships are your IP in some way.

Seth Marrs:

Yeah, hugely, right? Companies will hire you specifically for that, so why give it to… nowadays, the whole, it takes too much time, is kind of gone. Just turn on your recorder and off you go. And there is hope in this, that salespeople are starting to get enough value from doing these recordings that they are asking for it. It makes it easier for them, and even their customers are starting to ask for it, which is… Which is hopeful, at least. To at least have the context available. Now, you still gotta break all the walls down, but the context available is… is happening.

Jaya Gupta:

Right, right. And I think the other thing it’ll help with in sales is, too, just, like, audit… as you said, auditability and forecasting. Like, I think that… You know, right now, AEs or reps might say, hey, my deal’s here in the pipeline, but if you actually do some digging, you’re like, wait, no it’s not, but is there a way that, hey, based on all their unstructured communication, you could actually tell, like, have they actually CC’d finance? Have they actually CC’d, like, legal on the, stage? Like, could you update MedPig automatically based on, like, what’s, you know, the unstructured data that you’re seeing?

So yeah, I think that’s sort of the future for sales, and it’ll definitely at least lead to more, auditability.

Seth Marrs:

Yeah, I mean, I would argue that you would get a better result and a more accurate result if you did it that way, and just take the seller completely out of it. Their job is to beat the forecast that they give. If I have all your contacts, I don’t really need you to tell me anything, I can come up with a better forecast than you can.

Jaya Gupta:

Exactly. I think that’s very well said.

Seth Marrs:

So, one of the things I wanted to ask you is, because early days, like, not even early days, like, several years ago, the foundational thing that companies would do if they wanted to get all… everyone talked about 360 view of the… people still talk about 360 view of the customer, I want all of my data together, I want to be able to look at this stuff all the same way, and the way they did it was just dumping it all into a data warehouse, and like, look, I’ve got all my stuff together, so now I can have it and I could do all this stuff.

Never really worked. But what we’re seeing with AI is that almost doesn’t… doesn’t even seem to matter anymore. All I need is… all I need is MCP connections, and I… you can put your data wherever you want, keep it in the source of truth, and I’ll use those connections to be able to bring it together. Is there a need anymore for companies to spend all this time and effort to dump all their crap in, like, the same place to try to make it… Talk and unify?

Jaya Gupta:

Yeah, it’s a great question. I think that we’re… what I think that we’re probably in is a phase where everyone is experimenting with the tools. Everyone’s experimenting, okay, which are the MCPs that are actually useful? Like, people are experimenting with which models are useful. I think we’re probably in a period where everyone thinks that maybe they’re on… the rest of their life is co-work or codecs, but I think that that will… Yeah, exactly. I hope… I think that that will change. I mean, people spending too much time on Twitter probably think that that’s the… what the world looks like, but yeah, that’s gonna change.

It’s gonna be decomposed by… there’s gonna be hundreds of different models, like, and there’s gonna be a very, very different stack. Especially with how high the costs have become, and, you know, token maxing and all these things. And so, what I think will happen is that in the next, like, year and a half, everyone’s gonna experiment, and everyone’s gonna probably store things locally, or store things wherever, and let Claude take care of where to store it, or codecs, and not even worry about it. And then I think what’s gonna happen is, like.

Year and a half later, everyone’s gonna be like, shoot. like, what are… we need centralized organization, like, we need central information, like, everything’s a mess, and, like, I think what will happen is… For some companies, you definitely absolutely need data warehouses, and I would argue most companies do. I think that there are probably some companies that are using it that maybe realize that maybe they don’t need it.

Seth Marrs:

Yeah, so it’s almost… I mean, to a certain extent, put your data wherever you want, but at least know where it is, how it’s supposed to work structurally, so you can at least understand what’s going on, and it’s not good enough just to point a MCP connection to it and be at the behest of whatever that connection is set up to do.

Jaya Gupta:

Exactly, yeah. And I think while we experiment, I think, you know, everyone’s gonna lose a good year and a half or two years of, like, where, you know, where their data is, and, like, having central… and I think that’s also because people haven’t really captured or, like, unlocked, like, multiplayer AI in some ways.

Seth Marrs:

Tay more about that.

Jaya Gupta:

Like, I think the idea of, like… I think a lot of people, at least on Twitter, are talking about, like, how to have agents spawn off other agents and run thousands of parallel subtasks. But I think that what people haven’t really cracked is, like. with the AI and two humans working together, five humans working together, how does the AI best, like, kind of, like, sit in between, like, five different people, and, like, help and coordinate their work? I think Claw Tag was, like, their attempt at this. But of, like, you know, having a Slack integration and jumping into chats and, like, trying to progress things, but I don’t… I don’t think that that’s the right implementation.

It’s gonna look different. Of, like, how does the AI work ambiantly in, like, someone’s, like, group chat, maybe?

Seth Marrs:

Yeah, so people keep talking headless, and this, that, and the other, and the perception of headless is, like, I can just do everything wherever I want, or everything’s gonna be a chat interface or Slack. I think what you’re talking about there is it’s a little bit more elegant than that, and it’s going to be the surface area that you need at the point you need it. And you’ve got agents working with you in that surface area to progress you in the best possible way during that time. And nobody’s figured that part out, and there’s a lot of experimentation over the next year and a half that’s gonna have to happen.

Jaya Gupta:

Yeah, yeah, yeah, yeah, I think that’s spot on. I think it’s, like, whether it’s gonna be, like, some sort of shared workspace, I think Harvey is, like, starting to think about these things of, like, how do I have a vault that lets people communicate with one another? And, like, there’s obviously a lot of work that involves, like, sending information out to parties outside the company, and so, like, how does that, you know, the context graph for the external facing… Communications work. So I think that will be an interesting thing. And I think consumer is also gonna be an interesting, you know, side opportunity on, like, multiplayer AI.

Like, you can imagine, is AI gonna be in a group chat with all of our families?

Seth Marrs:

Yeah, and you’d think it was. How do you perceive that on the consumer side? Because the B2B side seems to be the easier play on that, and the complexity levels of doing it on the consumer side. You can almost see, like, an apple popping into that. Like, after everyone’s figured out all the hard stuff, they just come in because they have all that data and can provide a consumer product that’s way better than anyone else’s. Is that the right perception? Like, because it feels like consumer’s harder.

Jaya Gupta:

Yeah, I think consumer is a lot harder. I think that, like, I’m not sure if you saw, Meta, launching Muse yesterday, which is, like.

Seth Marrs:

there.

Jaya Gupta:

personal, kind of, like, AI app, and you can kind of, like, give it all your tasks, it’ll run… it’ll run longer… longer tasks for you. It’s… it’s very, very, you can integrate it with all your personal, kind of, like, you know, apps that you would use, Uber or Spotify, like… and so I do think that was the first thing that I saw that was, like, after ChatGPT, that was, like. For, sort of, like, single-player context for the consumer. And then there’s a company called Instinct, which has been making the rounds, if you’ve heard of it, that, you know, is, like, taking the aggressive side on iMessage of, you know, it’ll change your passwords for you to get something done.

Seth Marrs:

Oh, wow. Yeah, interesting. Very, very interesting. People are trying, they’re pushing that… That area, even though it is a complex and difficult area. Yeah. Okay, I… there… and I know you have kind of a bias here because some of the companies that you work with, but you talked about… like, you’ve talked about 3 different paths to… that people are taking to this. Like, you’ve got a build-from-scratch kind of agent-native product. like, it’s taking over workflow. You’ve got another one that is… just using traditional systems in place and kind of building workflows around it, and then the other one was around, like, a completely new system of record.

Like, you have a system of record, like ERP, but then a completely new system of record. At the end, which one do you think is… which approach do you think is best positioned to win, or is it kind of a mishmash, depending on the use case?

Jaya Gupta:

Yeah, I think it’s a mismatch depending on the use case and the size of the company. Like, I think it will be… I don’t think any… I can convince any CFO of any F500 company to about SAP, but I think it might be possible to convince them that, hey, you can overlay AI on top of some of the modules. And, you know, automate things like cash or clothes or some of the core accounting workflows without ripping out the GL. I think for maybe startups, they might be willing to rip out something like QuickBooks. And then I think that And sales, it’s also going to be different.

I don’t think I could also convince any F500, CRO to… to rip out Salesforce. So, I think it’ll just be very, very different, for each, category.

Seth Marrs:

Does it kind of just become same old, same old? You’re gonna have these companies kind of emerge that do those specific use cases really well, they’re gonna get very sophisticated with it, and then SAP’s gonna buy them, and then it becomes part of their… like, this whole kind of small to big push is just gonna… repeat itself. Like, it always has.

Jaya Gupta:

It always has, it always has, and it’ll maybe eventually create more mess for people, but that’s just what life is. But yeah, Palo Alto just bought Console, I think, and then, you know, you are… I think Salesforce is potentially buying Listen Labs, according to the news, so you’re… you’re gonna start seeing more activity pick up.

Seth Marrs:

Makes sense. Jaya, thanks so much. Really, really great to talk to you and hear your insights.

Jaya Gupta:

Thank you so much.

Julia Nimchinski:

Thank you, Jaya and Seth, amazing conversation. What’s the best way for the community to support you?

Jaya Gupta:

Twitter! I feel like, everything’s going on there, so feel free to follow me or DM me there.

Julia Nimchinski:

Awesome. Seth, how about yourself?

Seth Marrs:

Yep, reach out on LinkedIn, love to connect.

Julia Nimchinski:

Thank you.

Seth Marrs:

You guys.