How AI can help build smarter systems for every team with Eric Glyman and Karim Atiyeh of Ramp

In this episode of No Priors, hosts Sarah and Elad are joined by Ramp co-founders Eric Glyman and Karim Atiyeh of Ramp. The pair has been working to build one of the fastest growing fintechs since they were teenagers. This conversation focuses on how Ramp engineers have been building new systems to help every team from sales and marketing to product. They’re building best-in-class SaaS solutions just for internal use to make sure their company remains competitive. They also get into how AI will augment marketing and creative fields, the challenges of selling productivity, and how they’re using LLMs to create internal podcasts using sales calls to share what customers are saying with the whole team. Sign up for new podcasts every week. Email feedback to [email protected] Follow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @eglyman l @karimatiyeh Show Notes: (0:00) Introduction to Ramp (3:17) Working with startups (8:13) Ramp’s implementation of AI (14:10) Resourcing and staffing (17:20) Deciding when to build vs buy (21:20) Selling productivity (25:01) Risk mitigation when using AI (28:48) What the AI stack is missing (30:50) Marketing with AI (37:26) Designing a modern marketing team (40:00) Giving creative freedom to marketing teams (42:12) Augmenting bookkeeping (47:00) AI-generated podcasts

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:05Sarah Guo?

  Hi listeners and welcome to No Priors. Today, we're here with Eric and Kareem, the co-founders of Ramp, one of the fastest-growing fintech companies of all time. We'll talk about the path to self-driving money, how they make AI and systems thinking fundamental to every part of Ramp, and why using AI in financial services is not as risky as people might think. Welcome, Eric and Kareem.

• 0:28Speaker 01

  Well, today, we're super excited to have Eric and Kareem from Ramp joining us. So, thank you so much for coming.

• 0:32Speaker 02

  Thanks for having us. It's great to see you both.

• 0:34Speaker 01

  If you could tell us all about the origins of Ramp, how y'all got started, what you're focused on today, what roles you play at the company.

• 0:39Speaker 02

  Ramp, in some ways, is sort of the sequel or taking our last business forward. We'd started about a decade ago at a business called Paribus back in 2015. I would probably reframe it now as an AI agent company. Essentially, what it did is, let's say you bought something from Amazon, Best Buy, Macy's, whatever. Every store would guarantee that the next week the TV went on sale, you could get the difference back if you asked. So, we built it was an email app. We scanned your inbox for receipts. We detected what you bought. And if you were eligible for a price drop refund, we wrote an email as you, or chatted with the store as you, to ask for the difference. They would respond, give you the difference, and we charged a cut. We launched this. Within a year, we had almost a million members. And we had a life-saving offer from Capital One to buy the company.

• 1:23Speaker 02

  So, we were thinking a lot about, first, how do you turn data into savings? We didn't know the first thing about credit cards. We learned it. We saw it was a great, large industry, very profitable, but deeply misaligned with customers. People were obsessing over the question of how do you get people to spend more money or earn more points? And we got very interested in the question of, you know, if you actually listen to people that don't want points or cash back, they actually want more in their bank account. The best way to do it is not get people to spend. We get people to not put differently, like not spending $100 in the first place is 100 times better than getting a dollar or 1% back on it. And so, we got obsessed of, wow, what if you put Paribus in a card? What if you had a card and software that was designed to get people to spend less in the first place? That seems more aligned with customers,

• 2:04Speaker 02

  a different way to go to market, and fundamentally, a better product. And so... Back in March of 2019, we incorporated the company. Today, we're just shy of five and a half years old. And I would say a lot of what we're working towards today is, you know, this question of can you have, you know, self-driving money in an organization? Can you build really the primitives? You know, from one place, issue cards, make payments of all kinds, manage approvals, even automate accounting. But really what we're doing is when you have the right principles, abstracting away all the tedium. If you don't need to add receipts to a transaction, you tap a card, we pull the receipt from your email and you're done. You don't need to tag every transaction and how it should be booked in your books and records.

• 2:49Speaker 02

  And so today, 25,000 businesses use it from small startups to Shopify, to Boys and Girls Club of America, to farms, to anything in between. And, you know, so that's how we got into it, what we do. And I'm the CEO and run a lot of the, you know, business functions. Kareem is CTO, of course, technical, but also we get into it overseas. Even, you know, marketing, other functions that we see, think are becoming more technical.

• 3:15Sarah Guo?

  It's been an amazing run over five years. A lot of the value you're talking about are, like, as you said, for organizations. And remember, we're talking to you guys when you were just getting started. You were like, kind of discovering that there was all this value in businesses and how businesses spend. Instead of like, your first company was more consumer oriented. Like, how did you decide to, like, make that shift and, like, you know, learning about that audience, investing in it?

• 3:38Speaker 03

  I'd say a lot of the, like, really ideation phase of Ramp, we were talking to a lot of our friends, people in our community. And it just so happens that a lot of them were either starting early-stage companies or journeying early-stage companies. And a lot of the problems that we're facing at a larger scale were some of the ones that we're trying to solve for consumers first. And the funny thing with businesses is the better they got and the larger they got, the actually the more wasteful they would get and the less they would know about their spend.

• 4:12Speaker 02

  Tougher investors, yeah.

• 4:13Speaker 03

  Not only is it like a more interesting and in some cases like bigger problem to solve, it just like scales with success in some ways. So like the better companies were even more interesting opportunities for us. So we went after that. And there was, I'd say like another realization we had early around like you look at the user experiences of different products out there. And the ones we use as consumers on a day to day, like they obsess over every single interaction, every single flow, like Instagram's amazing. Robinhood did that very, very well for trading stocks and investing. And then those same people who use these apps in their daily lives show up at work and are expected to use tools that were built in the eighties

• 4:56Speaker 03

  and are incredibly slow and very clunky. And there just wasn't a good feedback loop between like what the people wanted to use at work and what they were using in their daily lives. And we saw an opportunity to really bring a lot of consumer thinking around like UI UX obsession to complicated business problems.

• 5:17Sarah Guo?

  Where did the first savings thesis come from? Like what, you know, if the idea is like spend less money and have less waste for consumers on like, there's a price change, like you go deal with this, that sounds amazing. Like what was the first step for a rep?

• 5:31Speaker 02

  First was just a basic business problem of, I think most startups, their biggest battle is people not giving a shit. Like you work on so hard on the software and there's a million other pieces, you know, apps, tools, cards. How are you going to stand out? And we realized there was this gap in the market. Most of the credit card ecosystem was really designed around kind of ego and excess and use my card and you get lounge access and points and you can fly to exotic places. And it just was so different from what business owners and CFOs and people I know who are trying to build enduring profitable businesses. So it felt very at odds. And so some of it was just, I thought there was a large unmet need.

• 6:10Sarah Guo?

  That's interesting. So just like the cultural premise of the card was very different. Yeah, it's not like, let's say like Amex fancy for the individual.

• 6:19Speaker 02

  The luxury that Amex, I believe, was really owning and building was almost like an 80s conception of like luxury of like you have like lounge access and it's covered in mahogany and it's super cool and like luxury in the 2020s is you know I can go to yoga class at 2pm because I have the time or I'm at the green juice because I'm healthy and and it was just very different and I think the thing that was so I think people were missing is time. People had very little time and so we tried to focus on not just saving money but saving time and to kind of zoom into where you know some of our first products was you'd see these great early stage startups really promising, starting

• 6:59Speaker 02

  to build traction and then you say, great, show me what you're spending on and you discover they're spending on like four sets of project management software or like five BI tools that do the same thing because someone tried it, someone passed the card to someone else, the finance person's like get me the receipts, doesn't ask the question of what are these these things and so some of the first insights was who you'd say, well, show us your credit card statement, we'll show you all the duplicative software, all the redundant spend and by the way, we just, you know, we helped you cut your spend by 2% per year. Maybe should use our card, we'll help you do that all the time. And so it was very simple was the mechanisms. It was it was cutting out redundant spend and it was automating processes. And

• 7:41Speaker 02

  I think it's expanded but sort of these same principles of you know if the one thing people don't have is an hour, you know, on Fridays to spend with their family because they're doing expense reports because they're tagging transactions, there's a tedious report, how can we give that back to them? How can we design products that will automate the tedious to, I would argue, gives people the new luxury of time?

• 8:03Speaker 01

  A lot of people say that with this wave of generative AI, there's new things you can automate or change and you all have been very big adopters of the technology internally in all sorts of different ways. Can you tell us a little bit more about how you're starting to use it both for internal purposes and you mentioned marketing is now a technical function, which is amazing. How you're starting to implement it for customers or where does that matter? How you think about that as a regulated entity? So I would just love to hear how you started thinking about using AI and then what that's led into for you all?

• 8:25Speaker 03

  100%. I mean, one of the early really thesis of Ramp is if we really want to help you save time and money, we need context, right? So one of the things we obsessed a lot over in the early days is We have some amount of information from the card statement. We have more information from the things that we see in your inbox. We can get even more information if we're connected to your ERP. And you get to a point where like, okay, there's a lot of it. It's very unstructured. How do you structure it and really help companies build and automate the workflows? So, and that's kind of how a lot of the like internal ways that AI shows up in our product really work. And they're like, they tend to, like, we really focus on the job to be done.

• 9:09Speaker 03

  So, and like, you wanna close your books. And there are different workflows that are part of this. And we're able to work on them a lot faster and really customize them without having to think about every company individually. Because we're able to just like apply a high level generic like AI algo with some constraints. And just make sure that those repetitive tasks become a lot faster. So, there's a lot of that that we do also on helping you figure out what bills to pay and when. There's often a right answer, right? You wanna pay the bill at the most optimal time. Generally, that's the last day that it's due. In some cases, it might be earlier because you get a discount. And those are the things that have a right answer.

• 9:51Speaker 03

  They take a little bit of thinking, but there's generally a way to evaluate and test what the right answer is. So, a lot of great applications in helping people just like get peace of mind on those decisions. And not have to like many look at every single invoice. There are many cases where financial operators stress over fraud, right? And the way that they check for fraud is they'll have to essentially look at different data sources and make sure that they match. That's also something that we could do like so much easier with AI. Like, let's make sure that we have three-way matching. And we can match your purchase order to your invoice to the goods that were actually delivered. So, there's many of them, but the one decision that we've made early is

• 10:35Speaker 03

  Financial professionals care about the job being done. And they care about the observability and having control.

• 10:44Speaker 03

  And a lot more than they care like they don't actually care that much if they're using AI. They just want it to be like fast and accurate.

• 10:50Speaker 01

  AI function is something that applies a lot of sort of thought and reasoning to different aspects of what built a pair other aspects of the business. And then my sense as well, you're also doing some really creative things internally.

• 11:00Speaker 02

  Yeah.

• 11:00Speaker 01

  Yes. But they can tell us a bit more about how that's impacting functions across the company today.

• 11:05Speaker 02

  A hundred percent. I mean, so a lot of people know outside in sort of said, Ramp is one of the fastest growing fintech of all time. One of the fastest growing software companies. Some of this is good positioning and timing and good markets. Some of this is like very, very early adoption of AI in into augmenting the capabilities of our sales team, of our marketing teams, of our underwriting teams all the, all the way throughout. And I'll, I'll zoom in on, on one. I think that there's a lot of startups now starting to think about,

• 11:33Speaker 02

  you know, AI sales and automation. And internally, years ago, we had built a functionally outbound automation team. And what we had noticed is this is back years ago, we had very little resources, but there was one sales rep who was incredible. You know, he could go out and book far more meetings than anyone else. And we were trying to figure out, we're like, wow, if only we had two of him. This would be great or three. And, and we want to understand like, what is making this person so productive? And so one of the unusual things we did is we had a few engineers sit down with him and just track what he did during the day. And he'd follow kind of his calendar and turn to wake up in the morning.

• 12:14Speaker 02

  And there'd be a new set of companies that raised funds or the people that he was in touch with who moved companies or all kinds of stuff. And he would, he would form his own list. Then he would go and try to guess at people's email. Then he would try to go and send different copy out. And it turns out that actually, so he had the right mechanism, but there was lots of manual steps. And so before jumping straight to, I'm going to hire a AI salesperson agent, we said, let's give him an Iron Man suit. Let's, uh, let's go and, and the things that he's looking at, let's go and pull those signals. So the lead list is done. For him, let's pull those emails. He doesn't actually have to guess at it. Let's run some A B tests. He can send things, but he's kind of based on this and figure out what's actually working and what's not. And the net effect is you kind of flash forward to today.

• 13:00Speaker 02

  The number of meetings that sales development reps at Ramp are able to book each month on their quota is multiples of any of our next closest competitors. And so it translates into of radically higher sales efficiency and the ability just to invest more heavily in in scaling. And so that's part of in how we've been able to continue to scale. You see similar aspects and cream can go deeper into how we're looking at it and aspects of marketing. But, you know, I think a lot of great marketing is thinking about, you know, a CDP customer data platform and understanding who people are, what's the intent and how do you generate great creative? In the world of AI that we live in now, the cost of creating creative art has never been lower in human history.

• 13:43Speaker 02

  You can create amazing visuals, images, texts, copy, you name it. And so a lot of what we try to do is sort of decompose what the function is and how do we use the radical improvements of foundation model capabilities connected to data.

• 13:56Sarah Guo?

  Well, one question on this, like, how do you, what do you think makes Ramp different in that? Maybe there's many things here, but, you know, most organizations, like the idea of resourcing, understanding the SDR function with engineers and then executing against bullying for them is just like, it's, it's not going to happen. Right? You know, division of the organization, we only have some engineers, engineers are not interested, like, what makes that work for you guys across the org?

• 14:25Speaker 02

  Two or three things. First, engineers are actually interested in and gold against business problems. Like actually drive a P and L, which is like very different often in many, especially like traditional companies. This is one of the things that felt like torture at Capital One of engineering and technology was a cost center. It was only an L. There was no P. There was no profit. And so it was just a question of where are costs and how can we rip things out? Not how can we grow revenue and profit? And that's fundamentally different from from the get go. I think it's important for any, you know, business to think about. I think going deeper, a lot of what people think about is how do we minimize costs? One of the inherent questions that we're asking is You know, time is money. Where are you spending your time? You kind of think back to the salesperson example.

• 15:08Speaker 02

  Really what we were most interested in is not how many dollars were we spending, but where were the hours going? And where could we automate? Which I think is a great framework to think about. Kind of the use of AI, which I think its best use case today is productivity. And so it shouldn't just be, hey, where is cost? And I would use AI to rip out costs. It should be, how do you augment and have every hour go a lot further? I think is some of the different framing. But you should add a couple more.

• 15:34Speaker 03

  A lot of it also goes to like the types of people we hire as well in the first place. I think a lot of organizations will have maybe a list of 10 competencies and skillsets and post the interview, you'll get in the room and it's like, well, this person checks eight boxes, but doesn't really check these two. And there are two very different things that we do is one, we really care about spikes a lot. And we also care about people who wanna, I don't know, are like very entrepreneurial and want to do things their own way and are kind of contrarian to some extent. And when you hire people like that, which again, some organizations won't hire because, well, I don't know if that person wants to stay here forever. Or I don't know if that person fits the Capital One mold, choose Capital One as an example, or really, really any other company.

• 16:18Speaker 03

  I think at Ramp, I really, I mean, from the beginning, I've always seen it as my job to just get like really raw talent that is incredibly ambitious. And it is my job to keep them interested and to keep the company interesting for them to continuously see these challenges and be attracted to them. So there are definitely different types of engineers that we hire in most companies.

• 16:39Speaker 01

  It's interesting because when I look at the companies that I feel that have done some of the most interesting things over time or have been capital efficient or whatever it is, they often end up focusing on certain forms of automation early that other people don't consider. So early Google was that way, actually. The online sales and operation teams, if you just extrapolated out how big it would be, by the time the company was 10,000 people, they would have had to add, I don't know, it was 50 or 100,000 people to just to deal with that volume. So they started building internal tooling early for it. So I think it's kind of a common theme for companies that are very thoughtful about this. How did you all think about what to build versus buy? Have you ever thought about actually spinning this out or offering it as a product? Customers, I'm sort of curious how you think about those dimensions of this.

• 17:19Speaker 03

  So I think that question gets asked a lot and like you generally get with a nuanced like it depends answer, which is generally right. But the one thing that doesn't get talked about is we can kind of assess generally whether if you decide to build, if you've done a good job building or bad job building. It's very hard to assess when you do, when you decide to buy, whether you did a good job buying or not buying. And it's not something that really comes up in, say, I don't know, a performance review or in the way like people get evaluated internally, which is kind of crazy when you think about it. Like we talk about people being great in organizations because they're great at hiring and recruiting. But you never hear anyone talk about, oh, this person like picks the right vendors.

• 17:59Speaker 01

  Yeah, I never measured up recruitment. That's super interesting.

• 18:02Speaker 03

  And that is one like skill set that we like, we did focus on very early. The way we pick the right vendors is almost like essentially interviewing them. We primarily like to talk to the engineering teams. We care a lot about the slope and the rate at which they're progressing as opposed to whether they have gaps today or not. And that has served us incredibly well.

• 18:26Speaker 01

  I'm curious, like, why not? And I know it's a very different type of business, so that may be the answer. But why don't you just start offering some of these services to your customers to use a SaaS products? In other words, there's this whole wave of AI happening right now. There's companies doing what you mentioned, SDR automation. The marketing stuff is a little bit behind, but of course you're trying to talk about it. Like you're basically dog fooding products. They could actually have real scale, potentially in the same customer segment that you have. I'm just curious, like, why not go and offer this to the world?

• 18:53Sarah Guo?

  Especially when like there are a lot of engineering teams out there who they experience these pumps pretty abstractly. Right? Like they've never, you know, been attached to an SDR for days on end like you described. So there are real advantages here.

• 19:06Speaker 02

  I think it's a fair question. I would say like the

• 19:10Speaker 02

  I know, it's like, should we need a strategy function? No, I want to come back in a week and be like, you were right. No, I mean, I would say the simplistic answer is just simplicity in the sale. Like, everything we do is centered around saving you money and time for finance organizations. And when you sort of think about like the order in which we've done things, it's, you know, expense and cards, instead of needing two apps to buy one thing and Amex and, you know, a Concur, you just tap a Ramp card and your expense report does itself. Then you add in bill payments and you add in procurement and travel. And so it's more products that continue to help the same economic buyer. And I think there's a lot of friction from a go-to-market perspective when you're selling to different buyer groups with inside of organization.

• 19:51Speaker 02

  And so I would say the overly simplistic answer has been probably we just haven't thought deeply enough about it. And maybe, maybe we should. I, I do think what you were saying though is, is actually, I think, worth emphasizing though. Like at a lot of companies, I think everyone's heard the saying of, you know, no one gets fired for hiring IBM.

• 20:09Speaker 03

  Maybe they should be.

• 20:11Speaker 02

  Yeah, I kind of think at Ramp, you might get fired for picking IBM, right? Like, you actually want to pick vendors that not based off of just they have every checkbox, but they are properly sloped and they're going to advance. And if you play this out in a year or two, they are going to give your organization a fundamental advantage. And kind of you think about just what AI can do of augmenting systems, like actually building great pipelines, rails by which to operate is, is, I think, never been more important.

• 20:40Speaker 03

  If you play it forward a little bit more, right? Like, if you have these like AI salespeople, AI finance people, AI, and how do you determine whether you picked the right one or not? If you're not, if you're like replacing a lot of functions that you would have hired like some people for with, with like essentially an AI professional? I don't think we have the frameworks or the tools to do that today.

• 21:04Speaker 01

  And also, how do you refresh that over time, right? Because often, once you buy something, there's this incumbency bias.

• 21:08Sarah Guo?

  Yeah.

• 21:09Speaker 01

  You just keep going with it. So you don't actually know what's in the market usually as well.

• 21:12Sarah Guo?

  One question on this, because like valuing time and productivity feels like such a core thing for RAMP. You know, a lot has to, but been on the board of companies that sell productivity in some way. And increasingly, AI companies are selling like productivity as part of the value. It's kind of hard, right? I would make an argument that a lot of organizations don't necessarily value productivity versus like of their people versus direct top or bottom line. And so I'm curious like how you guys think about it internally, how you sell productivity to your customers.

• 21:47Speaker 02

  We get classified a lot as a fintech company and like we're fine to be in that box. I actually think we're a productivity company. I agree. Like primarily what we're selling people is time. It's expense reports that do themselves. It's books that that do themselves, keep it clean or more accurate. It's a procurement like process that actually is just upload the contracts. We get the approvals. We show if there's better prices available. And you just have to worry about the thing you want to buy versus just jamming it through your organization. And I think it is productivity. And what I would say, some of what helps us is in the finance organization. Think of it from their perspective. Unlike R and D and the fancy parts of it, they're G and A. And CEOs say G and A has to go down every year.

• 22:31Speaker 02

  You cannot hire more people. And so they're asked to do the job of multiple finance teams as the companies get bigger without new resources. And we allow them to do that. We sort of create this the scale leverage for them to do that is is some of the way that we talk about it. It definitely helps to be a product that's free to try, that pays you to use it and shows you ways to cut out costs. And I think what Kareem was saying is it's funny. I think in 2024, we're almost like in this, I think back to the 90s, you know, in the early, probably 97, 98, your stock went way up. If you said we're whatever.com, it was really good. Then pets.com came out and there's a liability and it was really tough. And you know, now you've got saying we're AI this, AI whatever product like.

• 23:16Speaker 02

  We show, not tell is a big part of what we do. Like, we talked about automation, about increased accuracy because there's fewer processes that you need to go through. AI is there, it's some of how we do it, but it's not the way that you will meet with it. And so I guess it's all to say is like, we really focus on what is the outcome we're driving. How much will you save in terms of dollars? How many hours? And are you getting measured against this way? And can we really connect it, at least from a positioning perspective?

• 23:46Speaker 03

  I mean, on your point of like, how does that show up in the sale? It's like a third order effect, but like, we think that the companies that pick Ramp will become more successful, will grow to, in some cases, become bigger. And as a result, like, we will make more revenue, so.

• 24:00Sarah Guo?

  I love it. Natural belief in your own customer base.

• 24:02Speaker 03

  But I mean, there is, I mean, like, we're starting to have like statistical significance around things like, like, underwriting risk and, and like, risk of going bankrupt. And it's a lot lower for Ramp than like, anyone that we've seen out there. And we think, I'm sure some of it is like the types of customers that we attract. We think we have a brand that probably attracts people that do care about running a good business. So that helps, but also like, the hope is that like, if you're on Ramp, you're more likely to run a successful company.

• 24:30Sarah Guo?

  Okay, so you said you're a productivity company, not just a fintech company.

• 24:34Speaker 02

  Yes.

• 24:35Sarah Guo?

  But you are still a fintech company.

• 24:37Speaker 02

  Yes.

• 24:37Sarah Guo?

  And, you know, like, I talked to a bunch of, like, large customers, including like financial services, like traditional financial services players. And unfortunately, I would say, like, a level of adoption and use cases that matter is still like marginal in most of these businesses. And the reason I, you know, I'll hear from, let's say, large bank, Capital One-type company without naming them, is, you know, it's too risky, right? Like, AI doesn't work at the level of reliability on the workflows and use cases that we care about, and like, something, something regulators, compliance support, and I know you guys believe that, but, you know, you, I, like.

• 25:20Sarah Guo?

  I called Kareem and he's like, oh yeah, we're trying this and like it's, you know, it's useful in this way or we're building something internally. How do you guys figure out how to take that risk when you, you know, your processes, your product have to be like financial services, robust quality?

• 25:34Speaker 03

  I honestly don't think there's that much risk if you constrain the problem well, right? Cause at the end of the day, so like if you're, you're trying to figure out like, how to use AI to help you categorize transactions. It's not an open-ended question of like pick what, like make up categories. It's like, hey, I know what the right categories are. Like, tell me which one of these could it be? And the way this could show up in the product is every single form is pre-filled and you can edit it if you need to. Every single list is pre-ordered. And in many cases, we can be better at asking you just a question that matters as opposed to like, like the same form that we ask you every time, right? So like you might be traveling and you went and got a lunch. We don't have information to know if this was a lunch with a candidate or a lunch with a customer.

• 26:18Speaker 03

  We can ask you just that question as opposed to like, I don't know, that was how many people were at the event? Like, that's something that we can, we can, we can answer because we may have access to your calendar and things of that nature. So it really constrains the problem a lot more. And like, you're not just like having AI do completely unconstrained work, so.

• 26:39Speaker 01

  Yeah, I kind of categorize risks in businesses or at least ones that take risk from these sorts of directions as known risks and unknown risks. And you, you all have known risks. You know what you need to do, you know where things could go wrong, you can fix that. I think usually it's the unknown risk that causes real problems for companies. And so it's, you know, you have such a sweet spot in terms of what you build that I think that minimizes that, that sort of fear of something really bad happening, which is great.

• 27:02Sarah Guo?

  Well, I think this is like this ability to define tasks that make sense and understand like performance against the task internally is actually, I think, like a, like a very strong competitive advantage in application of AI today.

• 27:18Speaker 03

  And on the productivity point, it's like, we all, I think, can agree that like, AI is going incredibly good at translation, right? Like you could, there's some errors where like, okay, you're translating English to code and those could improve, but like broad translation, it's pretty good. The one type of translation problem that we see ourselves solving all the time is like, Accounting finance speak to like business.

• 27:38Speaker 01

  It's generically bad at math, too, right?

• 27:40Speaker 03

  Yes.

• 27:40Speaker 01

  Certain basic math just breaks, which often gets into financial related items.

• 27:44Speaker 03

  Fair enough, but there's not, I mean, there's, there's not a lot of that that we have to do, frankly.

• 27:50Speaker 01

  So can you, can you effectively, are you thinking of like fine tuning a model against certain accounting terms? Or doing, you know, like, I'm sort of curious, I think about problem solving, or is it just wait for future generations of models to come out?

• 28:01Speaker 03

  Or? We did spend some time fine tuning in many places and then we very quickly found out that our time was worth way more. And that we should just like wait, wait for like other generations of model. What we've gotten really good at, though, is hardening our infrastructure so that we can easily switch when we need to. And we can quickly evaluate the different models on the, on the subtasks that we care about. Like I was asked this question by, by, by one of our investors recently, it was like with like the GPT four O mini. How, how has that changed things for us? He has it brought costs down and how are we thinking about it? And my answer is like, oh yeah, it's, it's already in production. And for like 90% of tasks that we're running, it's good enough. So that was a quick switch. And within a day, we can know very quickly like that, yes, this is that good enough and we have the right evals.

• 28:43Speaker 01

  What is lacking right now from the AI stack? Like if you were waiting for one thing to happen, is there anything specific? Or is there a specific functionality or capability or just someplace where it tends to fail?

• 28:52Speaker 03

  I'd say like we, we have been exploring using AI to help you in, in essentially like end to end navigation of like the website or the app. And that, that's really hard.

• 29:05Sarah Guo?

  This is a cool demo like that I've seen, we'll, we'll link it in the podcast notes.

• 29:09Speaker 03

  Our app is changing so quickly and you need to figure out like where do you have like guidelines and constraints? And where do you let it be a little bit more open ended? There is a fun one that I, I really like, although it's not like quite there in production. And as Eric was mentioning earlier, like I'm really focused right now on like what lessons learned from really engineering and using AI in our product? Can we apply to other job families and processes that we run? And one of those is we, we, we write a lot of copy, we send emails to our customers. And at most companies, there's some kind of process where like someone needs to review the email to make sure that it follows your brand guidelines, that it has a clear CTA. And those are things that like in engineering, you could, it's more deterministic. You could do that as part like a test suite that you run during your like C I C D flow.

• 29:53Speaker 03

  What if we could build something like that for the copy that we're sending out? And we are starting to, to, to iterate on this. It's like every single email that is going out for the first time. Can you run it through like an A I review? That is a lot more close to instant and as deterministic as possible, although that's hard. There's definitely room for improvement there, but I would say the improvement is more around the interface and the knobs that you can use to tweak your model. Because today it feels like the most common way is just like write different prompts and longer prompts. And like that's, that's the main way that you can guide the models. I think Claude with artifacts, uh, it's something really, really different that, that, that I love. And it's maybe a new, um, I mean, it's a different interface.

• 30:36Speaker 03

  I mean, it's a really outputting a mini web app for you that you can tweak and, and change. And like we, we are obsessing over like what the right interfaces are for, for us.

• 30:45Sarah Guo?

  Maybe if you zoom out from that, um, idea of like, oh, like, well, we could have testing like we have in software and sharing, but on copy. Totally. To, uh, it's weird in general for a technology leader to own marketing. Like, you know, project out several years, what does marketing look like?

• 31:03Speaker 03

  Well, the, the one thing that I think will remain for a very long time is, is having good taste. And, uh, like at the end of the day, like if you are, even if you are working with A I's and, and machines, someone needs to decide like is, is this good or not? And, and there's this element of taste that you're not going to be able to replace. But what if you could get like the, the, the marketing teams and professionals to just really, really focus on that and remove a lot of the mundane repetitive. And that's really like what, what I'm obsessing about now is like, how do we give the people in our marketing more time to focus on the things that are true differentiators and not have to reinvent the wheel on, just really like processes that could be, um, hardened and, and improved with, with A I.

• 31:45Speaker 03

  And it starts with, there are different job families within marketing, probably a lot more than there are in engineering. The skill sets are very desperate. And very often, they need to work together effectively. Those interfaces between different teams are not always very clear. It's like the first step for me is really looking at marketing as any other system. It has bottlenecks, it has things that you could run in parallel, it has dependencies. And I guess the first step is like trying to identify where those bottlenecks are and building systems that reduce those bottlenecks.

• 32:21Sarah Guo?

  Ramps as systems architect, refactoring the stack and be like, I'm going to redraw this API. Actually, that's the problem, not the AI.

• 32:28Speaker 02

  So I want to tell like a, I guess like an anecdote for, so one of the things that's also unusual about Ramp is we're based in New York, very unusual for a fast-growing, tech-oriented company. And so we're on 23rd Street, but six blocks south of where we're located is Andy Warhol's old studio, which was called The Factory.

• 32:49Sarah Guo?

  Wait, I think where Eric's going to go with this is he's implying that because they're in New York, they have taste and that's the last remaining several people.

• 32:57Speaker 02

  There's people way cooler than us in New York, like we don't get credit. I want to talk about this, it's actually, I think, an interesting microcosm of, I think, where brand studios go in the future. So Andy Warhol's is a very interesting figure in the art world, which typically, take previous artists, you needed, if you wanted to paint the Mona Lisa, you needed to develop unbelievable skill over decades in order to finally make the thing, to sculpt something, to create. His big idea was, I'm just going to print stuff. And I'm going to get my production process so well-known and so well-focused that every day I'm going to print something out on silkscreen, use these commercial methods.

• 33:39Speaker 02

  And what I'm going to obsess about is not how do I make the art at all? I'm just going to obsess on what is striking. And if you kind of read interviews of what people would say at the time and what they did, it was this crazy place, they'd say every day, something new was how they did. They created lots of net new stuff, created reality TV. They talked about the security guard, Augusto, he was making. Art too and and people would joke about him that he would do this thing called art by telephone. And it was unclear whether or not Andy Warhol even made the painting and he would like sign it or someone would sign it as him. And you know what? It didn't matter. They were all Warhols. And you look at the collection of what he did over the time, like, there's no question. It was radically innovative, but you could tell it's a Warhol.

• 34:21Speaker 02

  There is some brand system going on. You can see it's distinctly him. There's a production and he abstracted all of the complexity of production so he could just focus on making the striking. And this was in the 60s and in the 70s. And you zoom out to now and it's like, gosh, like you can make anything. Like you can quite literally make your own version of the Mona Lisa. You can make music. You can make video. It's only going to get better. And I think it's an interesting lens to think about what might marketing look like in the future. Taste is very hard to replace. You might be able to see like, what do people click on or prefer? But, you know, all the complexity, if you really design and think about how do we produce thing?

• 35:03Speaker 02

  You can reduce down from where I think a lot of marketers and people making brand get really caught of. It's both, how do I create the environment by which I make the art? And then I test it and I'm focusing on the ideas and you see, I'm going to constrain all that. Just make striking, interesting stuff. And so I think one, building in different places, you think about different sorts of things. But I think that's part of why is like, when I think about, I don't know, I started talking about this. And I think when one of Kareem's, I think unbelievable strengths and superpowers is like, how do you create environment which you can create and give people leverage of any function? And I think marketing is this really interesting thing. It's one of the places that, you know, I think of most of the generative AI tools.

• 35:46Speaker 02

  A lot of it's around arts and poems and songs and music. And so I think it's a place people are really underestimating of the importance of what you can do to augment.

• 35:54Speaker 01

  How does that map against your marketing function today then? So I think, I think that's a really compelling vision of where things are going and how everything's evolving. I'm a little bit curious, it's like, how big is the team that you have right now focused on marketing? Has the use of AI or those technologies constrained how many people that you bring in? Does it give them enormous leverage? Is it, you divided in traditional like brand and digital and performance based marketing? And like, I'm just sort of curious, did it change the normal marketing department structure? Or is it roughly the same structure? We provide a lot of tooling.

• 36:22Speaker 03

  The structure that we have today is roughly the same and we're like primarily focused on giving them more leverage. I think what we're trying to do is like make sure that they're able to like match the speed at which we want to like continue to build product. And I think a lot of companies when faced with that will tend to slow down. It's great. We're going to stop shipping every week or shipping every month. Instead, we're going to do quarterly releases and yearly releases. And the problem with that is you kind of like cap yourself. And generally like you do that to give more breathing rooms or more to have more control over like the things you're putting out in the world.

• 37:04Speaker 03

  And the way we want to do that is by giving those teams more leverage and better tooling without compromising on speed.

• 37:09Speaker 02

  I do think the question you're asking is is completely right, though. It's like, what? What should any modern organization and function look like? And, you know, I think we're early midway. There's a lot of things that have actually been just like unbelievable in in starting to take a technical lens to functions that are

• 37:31Speaker 02

  sometimes underappreciated for how much technical work it takes to create great marketing, great sales, all of that. But I think the thing that is still hard for people to rock in some sense is, you know, if you have a great idea, you can you can spin up tens of thousands, hundreds of thousands of API calls and functionally have a team of a hundred doing some kind of tasks and just the human mind doesn't work in that way. And so I think part of what we're spending a lot of time talking about is is it a hard requirement where people need to think in systems and know how to build and use tools and thinking about whether it's a selection of vendors and great tools could be leveraged to even in your own working environment.

• 38:15Speaker 02

  How do you leverage yourself and employ that as a course thing we should be interviewing for everywhere?

• 38:21Speaker 01

  Like one of the areas at least I'm personally most excited about for AI. And in particular, if I look at the marketing use cases and I look at things like ads agencies, where they do a lot of the work that you mentioned in terms of iterating on copy, iterating on the imagery, doing it by different format for is it TikTok versus TV versus whatever? That is all very automatable with AI. And so I'm very excited about that area. It's just an area that's going to turn over because it's converting services into software. Large scale and a lot of what you're talking about is different forms of that internally.

• 38:48Sarah Guo?

  I do think you need to have this trait, like both in ICs and in leadership that you described of just like, like be able to picture scalability in a very different way, right? And I'll give you like one example, you know, a friend a while back, he owned marketing at Stitch Fix. Stitch Fix, like, you know, they have a bunch of stylists, they do some data science, they like, you know, communicate with customers and like most organizations, they've discovered video is a very effective marketing medium. And one of the things that a, the marketing leader there did was say like, okay, well, the traditional thing to do is to pay this agency that a lot describes to like, come up with something good and tasteful and on brand, whatever. And it's very expensive, the iteration cycle is a year.

• 39:32Sarah Guo?

  And he's like, well, I don't know, like we have paid all these stylists, there's hundreds of them. And like, can we just, like, we're not going to get the same level of quality. This is all pre-gen AI, whatever, right? But we're not going to get the same level of quality. What if we just give them direct manipulation and say, you all have to do video? It was a thousand times more effective and it was like essentially free because the like talent was in the building, even if the quality was not, it wasn't like Coke ad, ad agency quality as output, right? But what you care about is the outcome. And I think one of the ways I think about the creative fields and like, including ones that are commercial, like marketing, it's a commercial creative field is like, well, like, you know, if you, if you have people who think about scalability and you give people

• 40:13Sarah Guo?

  with taste and understanding of your business, like the ability to do direct manipulation, right? I make these videos myself or whatever, is probably going to be a lot more effective, but it breaks a lot of norms.

• 40:25Speaker 03

  That's exactly what we're doing, honestly. And just making sure that you give, well, you can teach a system what your design brand is and that system makes it very easy for anyone on the team to produce video if they want to, anyone on the team to draft the post of social content that is interesting. We start with what is, what works really well. And then from there, we iterate to make sure that it is on brand and can continue to improve and tell the like overarching story that, that, that we want. I think it's a lot more restrictive to start with, let's make something on brand and then let's make it work and a lot more costly. So trying to invert a little bit how we produce good creatives and good content.

• 41:09Speaker 01

  Is there any one of the last questions or topics that we should ask you about or they want to make sure to?

• 41:13Speaker 02

  We talk about whether it's interesting. So one of the things that occasionally gets talked about in the office, I think it was like a year or something ago, the Wall Street Journal had this like big article that came out that said, you know, there's a million fewer bookkeepers over the past decade and it's a crisis. And there's all these labor market problems. It's wages haven't gone up and so no one wants to be a bookkeeper and you're going to, you should expect inaccurate financials. And then someone, I think like a few months later, thought to ask the question of, you know, how many financial advisors and, you know, how many accountants are there? It's gone up by like a million. And I think it sort of speaks to the nature of jobs are changing and

• 41:51Speaker 02

  I think sometimes when people ask, is it going to automate everything? I mean, our view is like, it should definitely automate the tedious monotonicness. I don't think anyone should be spending any time you know, chasing people for receipts or dealing with the anxiety of you're the receipt person, you see someone at the water cooler, you can't have a normal relationship. That should be like, that should be like, you know, an automated system. That should be where Ramp is. And so I think if there's the jobs question, I tend to have a positive view. I think there is a place for taste and higher level work that people can do.

• 42:22Speaker 01

  So you're basically extrapolating out AI starts doing more and more things. And you're saying that means it frees people up from certain jobs that are just unpleasant, grindy, et cetera. It frees them up from a creative perspective because then suddenly you have centralized branding through AI and therefore anybody can start creating copy or using it in different leverage ways. And so you view it as a very freeing set of technologies?

• 42:43Speaker 02

  I think so. And some of this is for better or worse, just how I think about view the world tend to be, you know, quite positive and excited. But I I think it's true. Like when I, I think it's necessary when you just look at the increase of the size of the company, you know, hundreds of years ago, you didn't see organizations with thousands of people. Now there's, you know, million plus person organizations and to build this, you needed the development of just how do we measure the receipts and expenses at all these franchises? sizes in factories and kind of the rise of bookkeeping as a profession. But now is, you know, commerce is becoming digital in the first place. And receipts are automated and your books are kept for you.

• 43:24Speaker 02

  You don't need so many people and I think that's actually kind of a good thing because there's. There was this funny set of research we're doing internally and we wanted to figure out. So we have this function called strategic finance and we were interested just in like a benchmarking. How many people do strategic finance and it's something like, depending on how you measure it. I think four to nine percent of jobs in finance are strategic finance. Maybe this is where should we invest more, invest less, how do we, you know, rip out yield in the business? And I don't think this is finance people saying, you know, this is 91 to 96% to finance jobs are non-strategic. But sometimes it's sort of.

• 44:02Sarah Guo?

  Doesn't apply that.

• 44:03Speaker 02

  It doesn't apply it. It's a, it sort of feels that way.

• 44:05Sarah Guo?

  If you talk to strategic finance people, which you do, they're also like, well, like the part of my job, that's the most strategic part.

• 44:11Speaker 02

  It's very little.

• 44:12Sarah Guo?

  Really small, actually.

• 44:13Speaker 02

  But it's, it's real. Like I, I think part of what's made this work, one, I don't think it's possible to use ramp today without. AI is in your workflow in lots of places. You may or may not see it. It's there. It's part of why ramp is so easy and automated. But finance is somewhat unlike many other job functions and that people actually want automation. GA has to go down as a percentage. There's a lot of tedious tasks and what they want to be doing is saying, where should we be invest? They want salespeople not looking up people's email addresses, but going and selling with people, actually doing the high value thing. And so I, I do think we're actually, if we do things right, on a long run of actually having people work towards. Much higher value tasks and uses of their time versus just repeated things that can be automated.

• 44:57Speaker 02

  I, I think to make it work and to make it actualized, I think it's, I think there's something's got to give in terms of. You know, the foundation models are getting radically better, but you know, I was with the CFO of a company this morning who walked me through. You know, over $10 billion in revenue, but. Quite literally hundreds of finance tools in order to keep their books, make payments, receive payments. It's like Byzantine. It's crazy. The more to question, the fewer people ask is how do we actually build like the pipes and the raw primitives, you know, the card payments, the build payments, the actual accounting, the low level operational tasks and the piping that you need so that when you overlay intelligence, you not just get insights of what you can do and just go manage it across 100 systems.

• 45:40Speaker 02

  It's done. These card, you know, these vendors have been turned off. These ones have been turned on. This contract, we're going to renegotiate it and it's all orchestrated. And so I think actually thinking about the primitives, the orchestration, the way it works together is a necessary act to get the best out of the intelligence that's coming. So we can hopefully free people up, work on more interesting things, but we try to spend a lot of time thinking about that.

• 46:03Sarah Guo?

  Is it useful to ask like, what is exciting for Ramp that people don't know about yet?

• 46:07Speaker 03

  Auto-generated podcast, because that was interesting.

• 46:10Speaker 02

  Oh, yeah, yeah. It's very meta. It's coming for.

• 46:13Sarah Guo?

  Wait, no, no, no, no. There's taste here.

• 46:15Speaker 03

  Let's talk about this one, because it was fun and also very useful. But we have, I mean, at this point, we generate about like tens of thousands of hours of conversations with customers that we have all the time. You have lots of teams across the companies from engineering to product to design to marketing that would love to know like what our customers seeing, feeling, hearing. And it's very hard to do that because you can't listen to 10,000 hours. And we've had our like internal applied AI team that was able to put together a very quick process to essentially generate a five-minute podcast of, it could be Eric, it could be me, like asking questions of our customers. And then you get like five minutes of like the most interesting things that happened with customers during the week. We want to take that a little bit further so that anyone on the team could

• 46:59Speaker 03

  maybe zoom in a little bit on a subsegment of customers, a particular persona, a particular topic. But we think it makes like coordination. In the company a lot faster, so you don't have to like wait for information to make its way to you, you could just like go get it at the stores. Very interesting.

• 47:14Speaker 02

  My favorite thing is I think what the team wanted to show was, you know, it's the voice of the customer and we'd apply sentiment analysis to ten to thousands of calls to find like the happiest moments, people saying like, I love Ramp, it saved me, this is a very thankless part of my job, Ramp helped all this, it was fantastic. So that was like the emotional, that was like the podcast that like we wanted to make. And so we send this out, it's great, LLMs have time for it, in ways people don't. Start showing this to other founders and what every other founder starts asking us for is

• 47:42Sarah Guo?

  What's the bad?

• 47:42Speaker 02

  I want the voice of the angry customer, I want to know all the upset people.

• 47:46Sarah Guo?

  Yeah.

• 47:46Speaker 02

  Because as companies get bigger, I think the big problem is no one wants to tell you bad news.

• 47:50Sarah Guo?

  Yeah.

• 47:51Speaker 02

  People go, so having a large language model help make sure you know what's going on, unexpected use case.

• 47:57Sarah Guo?

  Yeah, very cool. Thanks for the conversation, guys.

• 48:00Speaker 03

  Thank you.

• 48:01Sarah Guo?

  Yeah. Find us on Twitter at NoPryersPod. Subscribe to our YouTube channel if you want to see our faces. Follow the show on Apple Podcasts, Spotify, or wherever you listen. That way you get a new episode every week. And sign up for emails or find transcripts for every episode at no-priors.com.