00:37:47 · George Bonaci
[00:37:47] S0: Yes. Yeah. Yeah. Yes to all of the above. Like, it's actually really cool being a user of Ramp as well as just working there. It's like every time I use it, it's it's just easier. It's more invisible to me. It's like, oh, I used to have to, like, take a picture of my receipts. Like, oh, now we're able to do some some on device, like, image recognition, and it'll pull the receipt picture out of my, like, camera roll for me. It's like, oh, I don't even need to do all of that. There's a back end of native integration. It's like, don't even need to do any of that. It... Like, that agent already looked at my calendar. It knows that this is going to be in policy even though it was taken on a... Or like the Uber trip was taken on a weekend. Like, knows this and knows that. And like, don't even see anything occurring. So it's happening in the background. And I think a lot more of of like the deep AI integrations in the products is going to mean that the work becomes invisible or the work that probably never should have happened in the first place. Like you now no longer responsible for. So it's it's very cool being on the receiving end of that. I do think though that the investments in the technology to make these things possible is is actually really, really important. Like a lot of the things that we end up providing to our customers or offering to our customers, like we built for ourselves internally. Like I spent a lot of time talking about how important, like being able to to route requests to the best model is because most people aren't gonna put that much thought into it. It's like, that's why we built a router. That's why we built like a universal harness. And they've produced such tremendous value for us internally. And this is like, at some some point, it's like, Well, why don't we just sell it? And I think that's part of the magic of Ramp, but it's also helped me deeply understand how the technology is is actually being useful even when you can't see it.
00:39:14 · Guillaume Moubeche
[00:39:14] S1: Interesting. To be honest, like, I love the the feature of being able to get it from the camera roll and not having to upload it. But a question for you on that specific topic because it's out of interest, I never truly understood why you needed to attach a photo. I mean, I understand, you know, like for an invoice or whatever, but let's say you go to a restaurant. In my opinion, it's pretty obvious that everyone knows that it's a restaurant and everyone knows that you're paying for food. You know? Like, so, like, how exactly... Like, why do you think no one has ever built something that's connected to pretty much, like, every single restaurant and that matches the credit card payment with, you know, like... And can kind of create the invoice for both and not having to... Because even the the fact of taking your camera or, like, your phone, like, take a photo of something, it's such a waste of time. If you think about it, you know, you have, like, billions of people on the planet. It takes maybe, like, five or ten seconds for people to do it. It's a huge amount of wasted time, you know, like, every day.
00:40:17 · George Bonaci, Guillaume Moubeche
[00:40:17] S0: Yeah. So... No. You're right. I mean, I I I think that's basically legacy work due to regulations and compliance requirements that that every company has just decided, oh, we just have to do this. You just have to take pictures of the receipt. But to answer your question, I mean, having deep integrations into these, the points of sale is is really like the unlock. But it's difficult because some points of sale, some restaurants are working on things that aren't internet connected or 15 years old or or or whatever it might be. So it's it's kind of like a a complicated distributed problem. But like our Uber integration is a great example. It's like we're we're integrated with Uber on the back end. I don't have to go and take screenshots of my Uber receipts. It's all just handled. And we're doing like more and more of that, but it is more of a complication due to the distributed technology used by all of the different things that you spend money on than than anything. [00:41:03] S1: Yeah. Makes sense. Talking a bit about competition, because I think Amex is is rebuilding their expense products around AI. What do you think ramp forced the incumbents to to change?
00:41:16 · George Bonaci
[00:41:16] S0: Yeah. I don't know. I I can't speak for them. But what I'd probably say is, I mean, we're generally always, like, more focused on the customer. This is, the most cliche answer, but it's, like, being more focused on the customer than the competition is is probably the key. Having said that, I think, like, you should respect your competition. You shouldn't underestimate them. You should see what you can learn from them, but you're always going to get significantly more return focusing on the customer than focusing on the competition. Competition. That's how you're going to get new ideas. That's how you're going to like think outside of the box. If you're constantly in the race of just trying to do what others are doing better or trying to beat them to something that they're working on, like you're you're not going to win the race. So that's my high level answer. I think at the same time, I generally think competition is really good. Again, very cliche answer. But, like, even internally in RAMP, I think, like, competition is good. If someone is ever just like, oh, I think I can do this better. Like, I... This person's encroaching on my scope or this is my job. And it's just like, if they think they can do it better, like, I don't know. I'll let them do it. Like, whoever can do it better can do... Like, does more of it. And I think that's true about, like, very, very zoomed in micro, like, oh, do not have enough work to do? It's like, alright. You guys are, like, competing on the same thing. Like, let's see who wins. That's also true, I think, like, macro when you zoom out. It's like, this is ultimately good for the customer if if more folks are competing.
00:42:24 · Guillaume Moubeche, George Bonaci
[00:42:24] S1: Yeah. I agree. And what would you say is like the the long term vision of of Ramp and where does the company wants to go in the in the next years? [00:42:33] S0: The vision has never changed, never wavered. I think that's actually largely, like what's driven a lot of the success of the company. It's always been to save companies time and money. I think a lot of the financial, like traditional financial products that we put out like very clearly do that. But now that more and more companies are spending money, not just on like program spend and head count, but now also like on tokens on intelligence, like that's that's a natural area that we can help companies save time and money. And I would even argue that it's it's a bigger area we can help companies save time and money because it's a very large... It's a very quickly growing cost center, but it's also one with, like, very little visibility. And I think that we have solved a lot of this for ourselves. So kinda like closing loop on the earlier part of our conversation is like, if we've solved it for ourselves and our mission is to help our customers save time and money, then, you should just give them what's working for us. [00:43:15] S1: No. It makes sense. And during your time at Ramp, like, if you had to do things differently, what are the things that you would change?
00:43:15 · Guillaume Moubeche, George Bonaci
[00:43:15] S1: No. It makes sense. And during your time at Ramp, like, if you had to do things differently, what are the things that you would change? [00:43:23] S0: I don't spend a lot of time thinking about the past. I think it's... You're you're always gonna get more returns thinking about the future and moving on from the past as as quickly as possible. But I think there's generally some things that I feel I could have, like, moved faster in terms of, resource allocation. I think that I was probably pretty slow getting on board with building, a growth brain of, like, like, a historical database and context of all the experiments that we've run. There was, a PM on my team that pushed for this relatively early. And I was of the page or of the of of the arguments. It's like, that's gonna take like two months to build. Like, ship the value today. Like, don't actually wanna talk about something two months from now because the technology is changing very quickly. She she was completely right and I was completely wrong on that. So there are like like mistakes like that that I could talk about, but it's almost like not worth it. Like, my advice to anyone listening is like focus on the future, not the past. [00:44:08] S1: Yeah. It makes sense. We talked at the beginning of the of the interview about, you know, like, the many bets that Ram could do and eventually, like, fail at because it's it's okay, like, to fail or miss. But what would you say is one bet that Ram cannot miss or cannot get wrong in the future?