ZipのAIスーパーエージェントが調達業務をエンドツーエンドで自動化する｜Rujul Zaparde

We can automate a lot of transactional work, but someone convincing another person about doing something and making initiative out of it and running a project, I don't see any AI doing that. I'm hearing a lot of reticence. I mean, everyone's kind of afraid of agents right now. Two of your customers are anthropic and open AI who certainly could build this themselves. It's very easy for enterprises to say, of course we're using AI, But the reality is organizations that deploy AI more deeply across their processes versus just, you know, they say they're piloting AI or whatever, see 39% ROI, if I remember, versus 6% ROI. Do you think about that at all, how the overall economy is going to change? Our plan when we started Zip was always to orchestrate the work that procurement sort of encapsulates so that we can understand the work, so that we can eventually help you do the work, right? And that sort of takes us to today and to super agents. And if you think about, you know, what the procurement process actually comprises of,

you could probably break down 80-ish percent of the procurement process into a series of relatively discrete sort of, you know, straightforward tasks, right? There's just a lot of transactional work. And I'll give you some examples. Like if you're doing business, if you're a company and you're doing business with a new supplier, you know, you're probably going to need to process a W-9 or W-8 type of document and make sure it's compliant. It's not more than three years out of date. You might need to if you're someone in your organization is procuring. They want to buy some new product, some new software product. eventually someone in IT will have to review a SOC 2 report, right, that they're going to route or stamp or say, wait a minute, there's like, you know, I think we've got to take another look at this thing. Either way, that's a relatively discrete task that can be automated, you know, with an agent. And so our original iteration of agents in Zip about a year, year and change ago, was to actually have, you know, many different little sort of discrete agents, if you will, that can be dropped at workflow that do all these discrete things.

But we realized that those agents were sort of one shot agents. So like a SOC 2 review agent, you know, we have a certain set of prompts. It would result in an output. And then the human, you know, you could have it autopilot or the human would, you know, be brought in to take a look and approve or not. But, you know, there's follow up that, you know, a lot of these things like require more back and forth. And so with super agents that we announced a few months ago, we prioritize really sort of discrete personas. So, for example, you can think of like, you know, the legal super agent in Zip can autonomously handle an entire end to end redlining process. It's connected to your email with MCP. It can send an email on your behalf to the supplier or to their legal counsel, process a redline, wait for them to come back, so on and so forth. And you can kind of like really take a lot of the human out of the work because AI, for the first time in this process, at least, handle more of the end to end work versus subcomponents of the work. And so that's that's something that we've been really excited about.

But you did mention, Craig, like, you know, procurement. You're right. Like I always describe it, you know, as every business in the world, by definition, has to procure things. Right. Like there's money coming in. It should be revenue. And then when you think about money going out, the top two buckets are always payroll and procurement. And in certain companies, like tech companies, most likely payroll comes first, procurement comes second. You look at more traditional industries, manufacturing, et cetera, almost certainly procurement will come first, payroll will come second. And then after a long, long, long gap, you have things like T&E and other things. But it's a very, very large area of spend for all businesses. Hi, the world is getting increasingly complicated. Every time your team deploys a new cloud workload or AI agent, another identity gets permanent access to your critical systems. Did you know 9 out of 10 organizations suffered an identity breach last year?

Because adversaries are logging in with valid credentials, not breaking in. Non-human identities outnumber human identities 109 to 1, creating a massive, unmanaged attack surface in cloud environments. Legacy tools were built to manage human employees, which leaves modern machine and AI access largely unmanaged. That's where Edira by Palo Alto Networks comes in. Human, machine, AI, one identity platform for all. 96% of human users operate with excessive permissions they don't actually need. Adira replaces permanent permissions with dynamic access so you can lock down every identity without slowing down your business. Adira provisions access on demand and revokes it immediately when the job is done, which removes high-risk permanent permissions.

With EDIRA, you get unified control by replacing siloed IAM, Identity and Access Management, PAM, Modern Privilege Access Management, and governance tools with a single control plane across human, machine, and AI identities. Secure every identity at your organization with EDIRA by Palo Alto Networks. Learn more at paloaltonetworks.com slash I-D-I-R-A. Well, when I said it's an interesting time, I'm referring to the rogue AI swarms that are broken out of sandboxes and things. And talking to corporate, you know, C-suite executives, I'm hearing a lot of reticence. I mean, everyone's kind of afraid of agents right now. I would think in some ways that works to your advantage because it's much more dangerous for a company to build its own agents to run around and do this stuff if they, you know, if they don't really understand safety and governance.

whereas presumably you guys spend a lot of time bolting down the edges on this. Can you talk about that, about safety and governance? Yeah, no, absolutely. I mean, you know, first of all, you're absolutely right. All use cases don't carry, within an organization, don't carry equal risk, all genetic use cases, of course. And what we find is, you know, procurement is one of the areas with the highest risk and the most sort of downside. Right. Like, you know, if you use an agent and it writes an email for marketing that sends it sends it out to someone and it wasn't Korea, it said something incorrect. Like they might complain, but, you know, whatever. It's OK. That's actually that is the right type of use case for AI to, you know, kind of go and be let loose a little bit. But when it comes to procurement and anything that touches the general ledger downstream in your ERP, right, and that is what's feeding your P&L, you get to an area of the world where there's no appetite for any.

It's about right. You know, none of that. Right. It's got to be it's got to be actually correct because you're putting, you know, this corporation at risk. You might have, for obvious reasons, you cannot have a P&L statement that is, you know, mostly correct or even off by even a sense, right? There's absolutely no tolerance for that. And so what we've spent so much time building is sort of this governed framework within which agents can live and operate and super, you know, our super agents, actually capability can live within that. And I'll give you some examples of that, right? And so like you can, you know, one of the key, key items that we have in Zip is you have to have, it goes without saying, full, full auditability of everything that an agent decides to do, does, the context that it, you know, wherever it's getting that context, it's like, you know, it's of course citing its sources. You can see when something happened. You can see, oh, this agent was actually allowed to, for example, that SOC 2 report example I gave you to say that, you know, this supplier has a sufficient SOC 2 according to our policies. who configured, you know, that capability and gave it that permission, right?

Even if it was done three months ago, all of that information lives in one place. And we make it really easy with audit packages and the like for auditors to internal, external auditors, able to download this stuff and validate it. So, like, that's a core, you know, kind of component of it. But what we also find is, like, you know, it's also really important to be able to say, hey, under these risk parameters, I always want, like, we work with a lot of banks, for example. I always want a human to review and sign off. And, you know, we have some tech customers that are more willing to say, no, under these parameters, I'm actually let it. I will let it run on autopilot and I will let it handle the situation and make it a call. But under these parameters, I still want a human. And you have that level of governance in the system. This becomes really critical because, you know, a system like Zip connects to your SAP or Oracle environment and actually pushes data. And we change records in that system. And so you really better be sure because that is the source of truth of the entire organization. And there's a lot of governance just simply has to go into that.

And so that's what, you know, we've spent so much time actually building this framework so that you can enable AI to actually take more work off of the human's plate. Yeah, and I read that two of your customers are Anthropic and OpenAI who certainly could build this themselves. So the fact that they use you guys is a pretty big endorsement. What models under – well, actually, on that point, how much do you benefit from their governance research? because there must be a dialogue about using this on their systems. Yeah, no, there absolutely is a dialogue. And, you know, what we've realized, like I'll give you an example that we solve for that is actually really complicated to handle. So say you have a user, right? Like we have our customers have pretty confidential data in Zip.

For example, right, if you're a customer of ours and you're going to go public and you've shared that with, you know, a consultancy that you work with, because you may need to share that type of information, that information might sit on an invoice. It might sit in a contract. It might sit in the statement of work because the IPO readiness work stream or something. Right now, that document is almost certainly going to be sitting somewhere in zip, going having gone through, you know, maybe a confidential request and approval process at some point. Now, if I as a user ask, hey, can you, you know, I have some question and it happens to touch that data, you, of course, don't want AI to use this other data to give me the answer to a question, you know, some other question that I asked when I can't, I should not be seeing that source. I should not, you know, know that information because I shouldn't have access to it, right? And so in Zip, we've built 350 different types of permissions, and those permissions apply to both humans and to agents. And the way it works is that if I, as the human, if an agent is acting on my behalf, that agent can only see the stuff that I can see.

So the human's permission sort of provides the boundary, and the agent can only reside within the subset of those permissions. And that's, one, quite challenging to build, but two, also equally challenging, if not more challenging, to actually test and validate and make sure it's working correctly, right? And, you know, there's just so much downside with this stuff. And so we've absolutely benefited a tremendous amount from both of those companies and their most recent thinking around governance. Not all areas of the enterprise are created equal. Yeah. Yeah. And is all of this sitting or based on one or two models? How many underlying models do you have running? and are they anthropic and open AI models or do you use other proprietary models or open source models in the architecture Yeah as you can imagine this is an area that rapidly changing is what I would say

but we're, so we continuously test different models for different use cases. So for example, we have customers that will process invoices in zip and coding an invoice turns out is a completely wildly different use case than having AI produce a red line on an MSA or having AI simply determine whether something is in the right, actually there's budget for something by trying to figure out what the right budget code might be and validating that type of logic or a series of other things. And so we end up, we actually look at, we do this on an effectively continuous basis, I would just say like a daily basis, where we actually test different models. We're testing all of the context that Zip has. And we actually, you know, also test with zip context and without to see what is the efficacy rate. And we track this like for many different types of use cases. So for invoice, it might be the line level coding. It might be the header, like we track header level, line level for, you know, purchase order.

You track all sorts of other stuff, you know, the clause level for contracts. So like we're always testing these things and we absolutely make tweaks under underneath the hood and behind the scenes to reroute depending on the use case and the complexity. Like, for example, we might have a customer with a, you know, 1200 page invoice, which might be hard to believe. But we have some customers that are buying tremendous amounts of individual goods from certain types of suppliers. And, you know, that whole invoice might not even be in English. Right. And so, like, that use case is very different than a one-page simple invoice. And so, you know, even that gets routed differently and tested differently. And so that's what we do. But, yes, we work with, you know, we use both of the models of the leading labs. And are you swapping in new models as they become more capable? A hundred percent. And what we find, what's interesting is that for certain use cases, like processing a simple invoice, even a more complex invoice, you don't need to actually use the most up-to-date.

You don't need Fable or you don't even need Opus for any of this stuff. Well, not any of this stuff, but for all the stuff, for like an invoice. And it's actually the context that matters way more. And so we help manage that, too, because, of course, there's a cost component to all the stuff. Yeah. Yeah. On the context window, there's when you say context, you're talking about what goes into the context window. Is that right? Or are you just talking about, you know, having the agents understand the context in which they're working? No, I think the former in this case. And yeah, I mean, it's an area that we really invested in because, you know, this is a high volume kind of area, especially. Yeah, and there's a lot of talk about the sort of quadratic expansion of the compute required as the context window grows.

And there's a lot of research into how to wrangle that or how to manage that. Is that an issue for you guys? I mean, if you have a 1,500-page invoice, a lot of models would get pretty confused if you put all that in the context window. I don't think it's been an issue for us. I'm also candidly not the expert in some of this just internally. I'll be open with you about that. But what I've seen, what we're doing, that has not been an issue. But, again, maybe I'm missing something. So what is your background and how did you get to Zip? And give a little bit of Zip's background. It's been around for a while. Yeah, absolutely. So, well, I guess my quick background, and I'll sort of segue that into Zip. So I dropped out of college in 2013, or, yeah, late 12, I guess, to start my first company.

Totally different business. It was called Flight Car. so we would give you free airport parking by renting out your car to someone else fully insured. If you remember, this was kind of the sharing economy, as it was called, you know, 13, 12, 13, 14 years ago, and Uber and Lyft and all the ride-sharing companies just recently started, like they were raising their stays and bees and that kind of thing. So people were, you know, it's hard to believe, but still driving to the airport consistently back then. And so that's how that's, you know, we started that company. Yeah. Some point late 12 ended up running the company for about four years. We it was a very sort of challenge, operationally challenging business to run. Like I quit making money in that business to like squeezing. It's like squeezing a lemon where the last drop that comes out of the lemon is the money you keep in a business like that. Of course, if you screw up halfway through, you don't need to squeeze the rest of the juice out of the lemon to know how things are going to end up, right?

And we had physical locations at airports. We operated at 17 different airports, and we had people running around, you know, washing cars and picking up folks from the airport. And all sorts of terrible things happened in that business, right? Cars get stolen. They get used for bad things. All sorts of other things happened. And we ended up doing an asset sale to Mercedes-Benz in 2016, which was not the outcome that we had set out to have. But, you know, I learned from that is, well, I don't think I want to do an operationally intensive business like this ever again. And the next company I start is going to be a VVSAS company. And that was at least, you know, that was all pre-AI and well before that, right, 2016. And so that was kind of the learning that I had. And then because I dropped out and not worked anywhere, I wanted to go work for a real tech company. And so I ended up working for Airbnb for a few years, which is where I met my co-founder, Lou, and had to go through the procurement process and became more familiar, I would say, with the pain points of procurement and all the stakeholders and, you know, just the complexity of it.

And then after that, I ended up working at YC as a visiting partner. I committed to being there for two batches. And then Lou and I had both decided that we would leave, be full-time on the company. but it still took us another four or five more months of just in the middle of COVID figuring out what we were going to do because we hadn't really had this idea yet. And then we started the company as it is today, August or so of 2020. So it's been six months. Wow. And you processed, I think I read, more than a trillion dollars in cumulative spend in that time. Yeah, I think it's like 1.2 trillion. But, yeah, that's how much has gone through Zip and been approved across all customers. When you say AI native, what's different about the way you've built your company that a Coupa or an SAP or Reba or somebody like that would have to do differently if they're building on a legacy system? That's a great question.

So I think there's a couple of components of it. Like, I think one, it's the people that you have, right? I mean, ultimately, our company is like the set of people that we have. And we make sure that product, engineering, and design, we only hire for those roles here in the Bay Area, in New York City, and in Toronto. And so we try to hire the best engineers, designers. Most of the engineers that we have, nearly all, actually come from a legacy-type background, if you will. They come from like more of a, you know, current generation. They work for a sort of tech company, a lot of consumer backgrounds, because we always say like, hey, we're solving for the end user. We're solving for the requester. If we make their life easier and we solve for their experience, it's going to help procurement. It's going to help legal. It's going to help IT. It's going to help all these cross-functional sort of stakeholders. And so having that DNA, I think, is maybe the hardest thing, actually, and the hardest thing to change. So that's sort of the first order of business for us. But the reality is we don't have, you know, 20, 30, 40, 50 years of technical just debt and, you know, extremely wide, sprawling surface area of use case.

I think that's what happens with some of the companies in the space. It just becomes, you know, you're servicing all these different industries. and we do too, but you have to be really disciplined and diligent about constraining the use cases and, you know, not building anything bespoke or custom fit for a customer, you know, in the old sort of world. Then you can't, you know, it can't be upgraded and you've got all these like sort of, you know, standout branches of the product. Like it just becomes really hard to take that and then turn that into something that like really utilizes AI. It's almost an impossible task. And thankfully, we don't have any of that. And I think we have got the right culture. It starts with the people first. You mentioned that your system handles 300-plus permissions. How are those set? I mean, if you're onboarding a customer, do they have to go through a toggle list of 300-plus permissions, or is there some way that that's automated?

It's a great question. So the way that the permissions work, so there's, there are, there, you can in-zip, you can create different types of roles. You can think of them as like personas, right? You know, ad requester, legal reviewer, so forth. So we come with a set of recommended default roles that are pre-configured, and then you can click into a role and see, okay, this permission's on, this one's off, this one's view only, this one's read, this one or edit, right? So on and so forth. And so those are also defaulted based on best practice. And frankly, just what we've learned and seen across, you know, you know, hundreds and hundreds and hundreds of customers. And so that's all configured for the customer. But what we find is, you know, a smaller organization might say, they might say, yeah, this seems fine. Like, we'll, you know, go forward with these. But, you know, sometimes you work with a, you know, Fortune 50 company and they want to, you know, they'll loop in their auditor and they'll actually, they will go through every single one because it's that important to them and that's what they want to do. And, of course, we're completely supportive of either approach,

but we want to make it easy for people and have defaults. And should they choose to go on at every single one, they absolutely can. That's how the permissions work. Yeah. In the – how much can the super agents do end-to-end without a human in the loop? Again, I imagine that's set by the customer, but are there certain things that you don't allow the agents to do without human intervention? Yeah, that's actually a great question. So a super agent can run an end-to-end process in Zip. So I'll give you an example. If someone puts in a request to buy XYZ type of good, you can have a super agent in there say, great we know you want this i'm going to go find a few suppliers that we already do business with and then maybe one or two that we don't do business with that would be good suppliers and run an end-to-end rfp process it'll actually find the suppliers whether they you know they're in your

vendor master which is the source of truth for your vendor suppliers in your erp it can of course run a web series go figure out like who would be best positioned to help you with this reach out to them, create an RFP, send them an RFP, wait for the results to come in, compute the results, make a judgment call or ask what it thinks the right sort of decision is based on cost and value, and then come back with a supplier selection. That's one example. Another example of an end-to-end process would be when something gets more progressed and worth the legal step to say, okay, great, I've got the customer's full legal playbook, of course, in hand, we've got the MSA. I'm going to actually annotate it, do a full, compute the full red line and send an email back to the supplier and let the supplier, you know, their counsel do their turn back and come back. And of course, after a certain number of turns or certain points remain unresolved, it will sort of raise up a hand and say hey I think I need a human lawyer to like get on the phone and you know talk to these guys But otherwise it can handle sort of that end process And so those are some examples And you know those are pretty complicated lengthy processes that exist

We've built the product so that, at first of all, just like, you know, when I use Claude or something else, it will check with me and say, hey, are you fine with this email? You want me to send it? And I can, I as a human can say, just don't ask me next time when I want to turn this thing on autopilot. and you can make that decision and it will run that end to end. Where there are more checks and balances in the system or is anytime we're pushing data into a source system, especially like an ERP system, when there's a sort of an integration push like that, we really strongly recommend there being a human in the loop sort of review just to make sure that certain types of records are correct because there can just be naturally more risk with that. But candidly in the enterprise, We see pretty much every customer wanting that. Well, if one of the agents does negotiate a bad term or approve spend it shouldn't have, who takes the hit? Is it the customer?

Is it zip? Or is it the model provider? It could have been a hallucination at the model level. It's a great question. The way I think about this is if I use Claude right now to send in, you know, I've got it open, of course, running something. But if I use it to send an email to you or someone and it's got a whole load of incorrect information, doesn't have the right tone or whatever, ultimately, and I gave it the permission to just go do this, it would, you know, that responsibility fits with me. I can't blame Claude because it just did what it could do. And I gave it sort of the permission and asked it to go do this thing. And so that's how I think about this, too, right? Ultimately, the responsibility does sit with someone. And it likely today in the framework of, I think, that is clearly evolving. I think it's this was a human that made the decision to have the agent do X, Y or Z and gave it the permission to do, you know, the X, Y or Z. And maybe that changes over time, but that's at least how I see it. I mean, it's fascinating that these agents, you were talking about sending out an RFP,

that the agent, does it create the RFP, the document, and does it send it to the vendors? Are those all autonomous actions? Yeah, correct. So the super agents come with default configurations, default prompts, right? So like, for example, if, you know, for a legal playbook, it'll have a default in there. For sourcing, so like putting out an RFX type of event, it'll have a default. Now, we work with customers. We've got a forward deploy engineering team that works with customers. Again, they can do this themselves, but many customers prefer that they just work with us and we kind of handle and configure some of the stuff for them. We can have the agent utilize a company's internal sourcing or RFX playbook for different categories of spend. And as companies get much larger, they end up having specific category managers that own certain domains like IT spend or certain types of spend in, you know, in cost of goods sold. Like we work with a large CPG company. And, you know, of course, they've got a category manager for sugar purchasing. Right. Versus cocoa versus whatever.

And so like we can those playbooks can be fed in in the zip for the sourcing, the super agents, RFX, super agent to to to to utilize. and yeah and then based on that absolutely it will it will create an RFP it will create an RFP and it will go send that thing out and it can handle all of this in practice what we find is you know your 1 million or 10 million whatever your thresholds are like the big right like your 50 million dollar AWS contract you're going to have a human do that just because it's so expensive it's like you know maybe your largest contract or whatever like right that stuff you're we find customers, they just want to have humans do it. But there's like a whole middle and then there's a long tail. And the long tail is not even being touched. The middle is not also really being touched because it's just too much work. If it's a $250,000 purchase or even a million dollar purchase from any organizations, it's just too much effort to do this. But if you have a super agent, well, why wouldn't you do it? And I think that's what our customers are seeing. And there's

real leverage there. I mean, these are dollars that are not, you know, sort of competitive, right? They're not like being put out to bid consistently. You know, companies are overpaying for stuff. They're not, you know, maybe getting the highest value sort of product or do it for their dollars. And AI can help with that. So, you know, there's real opportunity there. One of the things that fascinates me is, let's say that this is a case where the agent is drawing up the RFP and it sends it to a company. At this point or very soon, there's software on the other side that reads the RFP. And I was talking to somebody about FDA clinical trial submissions, you know, designing a clinical trial. You create a document that's six inches thick on all of these procedures and protocols, and they send it over to the – but it's an AI that is generating that document.

They send it to the FDA, and they just feed it into an AI. I mean, at some point, these agents will just be passing data. I mean, why go through the trouble of creating a human-readable document? Do you think that will ever happen on RFPs, for example? I think that's absolutely the world we're headed in. We internally use some AI tools for our own. We respond to hundreds of RFPs ourselves a year, and so, of course, we use AI as part of that process. It's actually interesting because I would say we had over relied on AI to actually produce some answers. And I realized that the quality of the output was I just, you know, I personally got dug into this maybe like eight, nine months ago, 10 months ago. And it was like some of the stuff is terrible. And so I actually, you know, forced our team to pull that stuff out and to force more human review of the content.

And the way we've now handled it is you can use AI like we have Golden Library of Answers that are human signed off on. And so then AI can take and match the answer to the question, but AI can't write the answer because it was writing all sorts of sloth. Again, you have to imagine it will get better, but, you know, quality really does matter because you don't want to be in a situation and respond in a certain way. And then the customer is like, well, you said this here. And we're like, you know. Yeah, my agent said that. Yeah. Yeah, that's amazing. In the materials that your team sent over, there's a UCI Health contract negotiation. Can you talk about that case? Yeah, yeah. So UCI Health, rather, is one of our customers. And I know that they ran – so they ended up utilizing a negotiation sort of agent use case with Zip.

And that drove a 20 plus million dollar sort of cost avoidance, you know, for this, you know, IT specific sort of initiative that they that they had. The reality is like so, you know, a company like UCI Health will have playbooks for how to run different types of negotiation. And it can be different types of, they're like different types of playbooks, right, for risk, for legal, for actual negotiation of, you know, the dollars and cents and the value and so on and so forth. And ultimately, what they realized is with the agent, they can simply, very related to the topic we were just on, cover more, cover more of their spend. And if you get to the middle, not even the tail, but like the middle, it adds up pretty quickly because even if you can shave, you know, 5%, 10% off on this one or 20% on this one or 2% on this other one, I mean, that all adds up. And it's very real, very, very real savings. And at the end of the day, I think you also feel better, right?

You're like, okay, well, I feel like I've actually tried and we've squeezed most of the juice that was left here in the middle of the tail. And so that's a big, big area of focus for us and customers. The money saved, I saw the number 20 million. That's cumulative. That's not in one contract. Cumulative, yeah, over a period of, I'd have to check whether it was like six months or a year or something like that. But yeah, it was over, you know, computed over a period of time. You have a report that surveyed over a thousand spend leaders. What finding in it most contradicted your assumptions about how AI is landing inside these organizations? Yeah, I think, you know, what surprised us was, and thinking back, like maybe it shouldn't have been a surprise, but it's the depth of the deployment that actually matters the most. Like it's very easy for, you know, an enterprise to say, oh, we're, of course, we're using AI, you know, we're piloting AI.

But the reality is that, you know, what we found in our data is that organizations that deploy AI more deeply across their processes versus just, you know, they say, you know, they're piloting AI or whatever. see 39% ROI, if I remember, versus 6% ROI. So like, you know, 6X or whatever, 6.5X, which again, should be no surprise really, right? Because ultimately, if you just have AI running and you're, you know, assisting a human with some basic tasks and you're piloting something, like you're not actually getting leverage because the human's probably still doing 95%, 99% of the work that they were already doing. And you actually haven't increased the surface area. You have to give the human more leverage, and you can only do that if you go deep in solving certain use cases. And, you know, what we've realized, too, is customers have access to agents. They have access to super agents and zip. They can turn them on. There's defaults in there. They can change the configuration if they choose and write their own skills and modify the skills, but they can just use super agents as they are.

But what we find is a lot of enterprises, you know, we've been we've created sort of this what we call AI spend automation offering. So we come to you as an enterprise and say, hey, look, we looked at your data. And in the last 12 months, you all have done, you know, 85,000 different types of approvals in Zoom. And we looked at what were you all approving? Right. We looked at all the different approvals and the steps. And we believe that we can work with you to create these, you know, 25, I'm making it up, solve for these 25 use cases with configured super agents, like configured to the bespoke use cases that you have. And that should result in, you know, an ROI for you of X, Y, or Z, right? Whether it be cycle time reduction, more cost savings, just because you're putting more requests out to bid. There may be some headcount savings with BPO, you know, type work that large enterprises do. And we will staff an AI FDE, an AI Forward Deployed Engineer, with you to actually help configure this stuff for you.

Again, they're not doing anything you can't do with a product. They don't touch code. But we find that, you know, we know how to, like, test agents and we can rewrite the prompt. It's all processing. We can do that for you. And that has, you know, really resonated. and we've got a bunch of customers actually doing these types of transformations with us because there's a last mile problem with AI and the enterprise. And so that, I think, was the main takeaway from or one of the main takeaways from this report. How do you see this changing commerce or the economy? Well, I guess before that, what kind of penetration do you have? Are there competitors? I mean, are you fighting for market share? Where are you? Is it sort of an open field and you just picking up as much of the market as you can as you grow How long before every enterprise is using something like this So Zip created the AI procurement orchestration category right which is sort of the capturing

the request with AI, handling the, of course, the routing of that request across all the cross-functional teams and systems that they sit in, like budget, risk, legal IT, and, you know, But today, much more increasingly so automating as much of that work as possible with AI. But we also do, we call that intake to procure. There's a lot of procurement lingo here involved. But we also do intake to procure to pay. And procure to pay is the traditional category of purchaser, you know, putting in a purchase requisition. It gets approved, approved, approved. It turns into a purchase order. Invoice comes in. The match happens. The invoice gets, you know, or the amount gets paid out to the supplier. that's sort of the incumbent category that's been around for 50 years or maybe even longer. And so Zip does that as well. We have our own product that does procure-to-pay in addition to procurement orchestration. And we're the only company in the world, actually, that does both. And we were fortunate to have been included in the Gartner Magic Quadrant for the incumbent procure-to-pay category for the first time as a visionary this year.

And so what I would say is that actually we find that in the enterprise, companies really want to move more and more to a platform. And they want a platform that is very AI first and has AI deeply embedded sort of throughout it. And when you look at the problem sort of that way, there's no other party in the world that does this. And I thought about why, by the way, too, because I've been like, well, why is that? And I think the reality is some of this stuff is so complicated and requires there's no margin for error. Right. Like we have to handle, you know, tax requirements in the country of Brazil. because we have many customers that do business in Brazil and the tax has to get handled correctly. Obviously, it can't be wrong because if it's over, you just overpaid. If it's under, you might have a regulatory problem. And so it's got to be correct, right? And that's just like one example. And so it's all of the permissioning, all of that depth is really, really hard to do. And so we've been continuing to be focused on just building more depth into the product and more depth into the AI components of zip end to end

and then capturing the end-to-end opportunity. In the market overall, I mean, is it sort of an endless market out there? Or where do you see this growing? You know, I think what's been interesting is, you know, it's so horizontal, the problem that we solve. Like we've got, you know, scientists at some of the leading pharma companies that utilize Zip to buy chemicals for the drugs that they're, you know, testing and creating. To you have researchers at the labs using Zip to procure whatever they need in their day to day jobs. To you have, I'm trying to give a good example, like, you know, a company that makes chocolate bars procuring everything that goes into the chocolate bar. through Zip, right? So it's just such a wide breadth of sort of use cases that we have. So, you know, in a sense, it truly is horizontal and it truly actually is kind of endless.

It certainly feels that way because every company in the world, at least above a certain size, maybe a few hundred employees at the absolute low end, has this problem. It's just so universal, right? Every company has to procure things. And as economies progress and there's more specialization in the economy, companies end up procuring more things because you can just focus on the thing that is your special sauce and your comparative advantage. And why should you, you know, why wouldn't you procure all the other stuff? Because it doesn't make your business special. You can focus on the thing that does. Right. And, you know, just we're at a moment in time, right? I have to keep reminding myself this. So something like Zip in 10 or 20 years, whether or not it's Zip or that process, that software layer, will have grown both in the market and, as you say, in its depth.

Are we, you know, Yuval Hariri talks about eventually the global economy resting on this fabric of agents that do all this stuff. Do you think about that? Do you see that coming? Do you think that's a sort of misguided prediction? and what do you think the implications are, again, for having these unseen AI agents taking care of all of this routine work? I do subscribe to it, and, you know, for the most part, like, we will, and I absolutely believe that we will have our customers just automating more and more of this workflow and the reviews that the procurement process is comprised of with AI. So much of that work is fairly transactional. It's important, but it's transactional.

And you can see more and more of that getting automated. But what that implies, though, is that the governance around this and the framework for that governance, the permissioning, the auditability, the controls, only becomes more and more important, right? One thing I didn't mention, but the majority of our customers have Zip in SOX scope. You know, SOX is an extremely critical framework, obviously came out of the Enron scandal in the early 2000s. Like, you know, you have no margin for error on Sarbanes-Oxley, right? And so it really makes the framework more and more crucial if you're going to give more rope to AI. To give more rope, you have to have the framework and the controls around it. And so I think the framework, the controls become more important. And I actually think that, you know, what I'm excited for is that all of us, the humans that are hopefully left in this world will actually be able to focus on more exciting work. Like, yes, a lot of folks say strategic work, and I totally agree with that. But I also would like to believe that that type of work is actually more rewarding.

It's more fun, and it's more fulfilling than just a lot of the transactional work I see. Companies, employees of companies, organs. It's also going to have an impact on the economy in that it makes pricing much more efficient. But a lot of people live off that inefficiency. You know, a lot of businesses, you know, capture that overspend because the system is inefficient, because they're friends with the buyer, the buyer is friends with the vendor, and they sort of don't press too hard. hard. I mean, that's interesting, too, how over time the economy will become increasingly

efficient. But at the same time, a lot of the rent takers, as they're called, will eventually lose their advantage or their edge. I mean, do you think about that at all, how the overall our economy is going to change. I do subscribe to that. I totally agree. And I think it's a good thing for the world, right? I think, you know, some of that inefficiency in the sale process, whether it be because, you know, I know a guy or because, you know, products, certain products, even like ours need to be sold and you need enterprise sellers. And, you know, ultimately that is the whole process is, you know, fairly inefficient and political and this and that. But you have to imagine that if you if you squeeze more out of that and there's more value for organizations around the world, they can take those funds, they can take those savings, and they will, and the economy will reward those who then reinvest those funds further into research and development, making their products better.

And ultimately, that's going to likely result in hiring more people for other jobs and other roles. And, you know, whatever these types of macro changes have occurred in the economy, you know, it seems like things just get moved around and there are actually more opportunities that can create it. So I'm certainly, you know, an optimist when it comes to the implications of AI. I would also say, though, I was reading something somewhere that said, you know, it was like the last job that will be replaced is that of an enterprise seller. And, you know, unfortunately, I agree with that because, you know, we can automate a lot of transactional work, but someone convincing another person about doing something and making an initiative out of it and running a project. And, you know, I don't see any AI doing that. Not until the super intelligence arrives when they'll be able to convince any of us of anything. Yeah, but it is fascinating. I mean, you're in this new category.

And as you say, it's horizontal and seemingly endless how much market you could capture. But at the same time, you're part of this global transformation that we're really only at the very beginning of. And you're part of it. I mean, it's really interesting. So what's next for super agents for Zip? We're excited about is continuing to work with customers to build more and more sort of super agents that can take on more and more personas. Like we prioritize the first five. We intend to prioritize the next 10. And then layering that in. I mean, we're very, very much hiring more forward to put engineers. That's something we can't have been a little bit behind on to work with customers to transform their processes end to end with AI. And I've been excited about the work that we're doing with that,

Tagovic Transformation, across all industries because we've also, and I've also been learning a tremendous amount. But what an exciting time. I agree with you. I mean, it's actually only been four years that we're into this transformative change in the world, almost four years. And while it feels like a long time right now, I'm sure in 10 years, there'll be so much more that will happen. That's right. And I have to ask, so you and your partner have the idea. It's hard to hire people to build stuff like this. Where did you get your team? Are you guys both coders, engineers, developers to the point where you could build a proof of concept on your own? Of course, we've hired a lot over the last six years. Yeah, I mean at the beginning, yeah. Yeah, yeah, yeah. At the beginning, we really actually brought on a lot of or a number of really close, awesome engineers and designers and PMs that we work with at Airbnb and before people that we just knew really well and were like, we know they're awesome.

They also very much needed to take a bet, too, because we had pretty much nothing happening at the time, right? No customers, basically no product. And so, you know, I'm grateful that they took the leap of faith and chose to work with us way back when. But that's how we got going. It was people that we trusted who we knew were, you know, absolutely TEDx engineers. Okay, I'm running out of stuff to ask. Is there anything I haven't covered that you want listeners to hear? No, I think, I thought that was great. Yeah, thanks, Greg. Yeah, I was saying to Krista before, Rujul, that that's one thing I love about this. This is an area that I've never thought about. But when you start thinking about it, yeah, it's interesting and how it's, yeah. I didn't either start a company, so, yeah. you