Simplifying Software Integrations with Merge’s Shensi Ding

Episode Overview

​Shensi Ding is the co-founder and CEO of Merge, a pioneering company in the field of unified API integrations. Before establishing Merge, Shensi gained valuable experience in the technology and finance sectors. She began her career as a Computer Science Researcher at Columbia University, where she earned her Bachelor of Science in Computer Science. Following her research role, she worked as an Investment Banking Analyst at Credit Suisse from July 2015 to June 2017. She then served as an Investment Associate at Silver Lake from July 2017 to April 2018, and subsequently as Chief of Staff at Expanse from April 2018 to May 2020. These roles provided her with a diverse skill set and deep insights into the challenges of software integrations.

In 2020, recognizing a significant gap in the market for efficient and scalable software integrations, Shensi co-founded Merge with Gil Feig. Their vision was to simplify the complex landscape of B2B integrations by offering a unified API solution. Under Shensi’s leadership, Merge has experienced rapid growth, accumulating more than 3,000 customers, including notable companies like Ramp and TripActions. The company has also secured $75 million in funding from prominent investors such as Accel, NEA, and Addition.

Shensi’s leadership style emphasizes team cohesion and strategic thinking. She is known for her dynamic approach to problem-solving and her commitment to building a strong organizational culture. Her forward-thinking mindset and dedication to excellence have been instrumental in positioning Merge as a frontrunner in the tech industry. Beyond her professional endeavors, Shensi is passionate about mentoring aspiring entrepreneurs and contributing to the broader tech community.

Episode Transcript

 Kevin Rosenqvist: What about data security? With more and more, you know, systems getting connected through APIs? What do you think the biggest risks they face are, and how can the industry address them?

Shensi Ding: I think the future risk that will probably happen is people will authenticate their integrations and connect them, and they might leave the company and the and the head of security or CISO just has no idea that that integration exists, that to know that those credentials are there, to even know that the data is flowing somewhere else, and maybe you churn from a vendor and but they have like live API keys. They’re consistently making API requests. I think that is something that people are going to start thinking about more in the future. We’re still actually in the early days when it comes to integrations, so I don’t think it’s a problem yet. And usually you require admin access in order to do that in order to get the credentials and share them. But I think long term in the future, especially as individual authentication becomes more prevalent, as more use cases come like start to rise, it’s going to be a little bit scarier for a lot of security teams.

 Kevin Rosenqvist: Hey there and welcome to PayPad. Where we bring you conversations with the trailblazers shaping the future of payments and fintech. My name is Kevin Rosenqvist and thanks for being here. Software integrations are the backbone of modern businesses, but they’re often more complicated than they need to be. Today, I’m joined by  Shensi Ding, co-founder of merge, a company tackling this challenge head on with their unified API. We explore the broader world of API integration, the evolving role of connectivity in tech and entrepreneurship. I also get her take on where technology is heading and if those who fear it need to pump the brakes a little bit. Joining me now from New York City,  Shensi Ding, from computer science researcher to co-founder of a fintech company. And you, you did it in just over five years from what I could see on LinkedIn. Did you always have an entrepreneurial spirit?

Shensi Ding: Well, I yeah, I guess I did, yeah.

 Kevin Rosenqvist: According to hey, as long as you didn’t lie on your timeline. I mean.

Shensi Ding: Uh. Well, yeah, I never actually thought about that, but. Yeah, so not really. I think it’s something that’s always fun to talk about. When you study computer science and you’re in college, you have a lot of friends, and you’re just like, hacking. It’s something that you always talk about for fun. But it wasn’t something I seriously thought about, especially since I went into finance after college instead of being a software engineer. And so when you do that, it becomes a little bit harder to be a founder.

 Kevin Rosenqvist: Yeah, yeah, yeah. What’s been the biggest challenge so far in transitioning into that entrepreneurial founder role?

Shensi Ding : I mean, every day is just so traumatic. And I think you don’t really know. You really don’t know how traumatic you think every job is, like really hard. And then you have this job and it’s just it’s fully consuming it, like it’s all you think about. It’s all you do. You are also like, all these people are very dependent on you. And like the success of the business and the team that you’ve put together. So I think it’s just how all consuming it is. You don’t really think about that. And even when I did investment banking and I was on call all the time, It’s nothing compared to this because there’s just so much more pressure, and you can actually clock off when no one’s messaging you in banking versus in this, like you’re never really off.

 Kevin Rosenqvist: Do you enjoy it?

Shensi Ding: I love it. Yeah. It’s fun. And like I do a little bit of everything and I’m not really good at one thing other than annoying my team and like, making them do a great, great job. And so what’s really fun about this is like, I can, like, dip my toes in everything. I can do a little bit of marketing, I can do a little bit of, I can code still if I want to, I do sales, I do customer success. And so being able to do all of that is really fun for me. And I also just it’s I’m able to build the team and the company that I always would have wanted to work at, which is also very rewarding.

 Kevin Rosenqvist: Yeah, that’s for sure. Like, this team is everything. I’ve talked to a lot of entrepreneurs here, and pretty much everybody says it’s really they’re only as good as the team that they hire.

 Shensi Ding: 100%, yeah.

 Kevin Rosenqvist: Do you find it hard to find good people? What do you look for generally?

Shensi Ding: So my hot take is if you think it’s hard to find good people, you’re never going to find the people. I mean, obviously, like, there aren’t that many good people, right? Like, but if you think that you are not able to hire good people, then you only get shitty people at your company and you need to really keep your standards high and you can’t lower them just because it’s hard. So, for example, like when I was looking for my CRO, I think I outbound it like hundreds of people personally like with messages and then also over email and then also on LinkedIn. And I would like try and also use the search firm. And it was very very difficult to find him. And we only gave one offer like it was very difficult. And I probably went on almost 100 calls trying to find this person. And it’s very easy to be like, okay, like this is the best I can get. I’m just gonna settle with this. But you’re always going to fire that person because it’s not a good fit. When you really end up fighting somebody, you’re like, wow, this person is going to be a game changer for my company and my team, and I really think it’s a good fit. And they come. Oh, that’s such a good feeling. And especially when you’re right, that’s even an even better feeling. So yeah. Yeah.

 Kevin Rosenqvist: So I’ve been wrong many times before. That’s not a great feeling.

Shensi Ding: Yeah.

 Kevin Rosenqvist: Are you more of a do you look for certain skills or certain personality traits. Are you more of a gut feeling kind of a person when you’re hiring.

Shensi Ding: It’s definitely part gut, but I think it’s because I’ve had so many data points of interviews that at this point, I can just tell if it’s going to be a good fit or not. I can tell the type of responses people give me, like they actually want this job or not. And if they’re like, oh, like they didn’t do any research, they don’t even know, like they don’t know what the company does. They say all this like stupid stuff or they like backtrack. And then I double click on, like, their previous jobs, they never cared about any of their previous jobs. And I just know they’re not going to care about us either. So I can kind of tell in the first 5 to 10 minutes, but like obviously you can you need to double click a lot more and spend much more time with them to actually know if it’s a good fit. But yeah, it’s part gut, but I think it’s because I’ve had so much practice at this point.

 Kevin Rosenqvist: Yeah, I feel like a lot of this these days, because it’s so easy to just fire off resumes that people just say it’s a volume game and they don’t really like you said. So people sit down, they don’t even know that much about your company, and you’re sitting there talking to them about a role. It’s like, do you really think you’re going to get a job like this if you don’t really research the company and understand what they do? Yeah, I know, it’s sad. Yeah. So let’s talk about, uh, some software integration, shall we? Yeah. Software integrations have been a critical, messy part of tech for years. Why do you think integrations have remained so challenging? And how do you see the landscape having evolved recently?

Shensi Ding: Up until this point, it really was like each integration was a special snowflake, and it was really hard to get the resources to build those integrations in-house. What was happening in like the 20 tens and like early 2020s was you’d buy two pieces of software and they wouldn’t have an integration with each other. So you would need to buy a third piece of software that would allow you to connect the two. And that process would be really difficult because no one, no one really wants to do it. It’s like kind of hard to get someone who’s technical enough to do it internally. It can take a lot of time and those integrations could break. And so it just wasn’t something that people enjoyed doing. And you don’t get a ton of revenue out of it. Like the most that you really get out of that process is, um, cutting expenses potentially or, or some form of automation. But now, increasingly, the buyer just has so many options for what vendor that they purchase. They just expect the vendor to have those integrations out of the box. They don’t have to buy that third piece of software, but the ability for that vendor to have the number of engineers and to have the partnerships and to be able to build and maintain those integrations is so difficult that it’s not really possible for that vendor to satisfy all the customer’s demands.

Shensi Ding: And it also is a huge time constraint, because you’re essentially limited by the number of integrations that you have. If you’re trying to sell to a vendor, sorry, if you’re trying to sell to a prospect that is using a vendor that you don’t integrate with, you have to disqualify that. There’s just no way they are going to buy your product. And so I think it’s just gotten very, very messy and out of control. And it’s just something that there actually aren’t that many people who really specialize in it in the world. And I think, like when I first entered the space, I was like, oh, there’s a lot of engineers, they can do it, but a lot of people are doing it for the first time, and they might do it incorrectly the first time. And so that just creates a lot of tech debt. And so the second person that ends up going on to it sees this, like Frankenstein code ends up having to modify it even further, and it just creates a mess. And so overall, like the integrations industry is very messy, both internal and also external facing integrations. And there’s been a lot of different ways where people have tried to solve it, but nothing really perfect up until hopefully merged.

 Kevin Rosenqvist: Yeah. Yeah. And what was there an aha moment for you as far as being, you know, founding merge, was there something you were like, man this is something that we need to we need to do.

Shensi Ding: So not really in the beginning actually we, we knew that this was a problem we wanted to solve. So instead of solving the internal integrations problem, we wanted to solve the product integrations problem. And there wasn’t anything that was really solving it. And we thought of a lot of ideas that were pretty similar to the existing solutions, and they just weren’t great. And when we thought of our first new solution, which was like a one to many API or what we call a unified API, we actually weren’t sure if it was a great idea. And we talked to a lot of companies trying to get feedback on if they would pay for this. And everyone was like, if this works, we will pay for it, but it’s going to be really hard to make this work. And we agreed it would be really hard and it is really hard. But I think when we actually had our moment was a year in. Our product was honestly terrible, like it barely worked. It was really horrible, Like, one customer loved us and another customer hated us. The customer that hated us was saying like they were like, you need to fix this or we’re going to churn. And they actually just never turned because they needed our product. And so that motivated us for thinking like, wow, we actually do have product market fit because everything is breaking. We hate our lives, but this customer is really actually using this product. And they’re yelling at us because they’re using this product and.

 Kevin Rosenqvist: They want it. Yeah. If it was easy to get rid of, they’d just be like, hey, you know, get lost. You know?

 Shensi Ding: Yeah. And they wanted us to be successful. And so we tried to make it successful because they wanted it. And it was much better to have someone who was yelling at us than someone who just didn’t care at all. So I think that was actually our first aha moment, was realizing that, like, people cared enough to yell at us and like, threatened to churn. And that’s not true because they needed the product.

 Kevin Rosenqvist: Yeah, that’s a good feeling. That’s a good feeling. I mean, it’s never good to get yelled at, but it is a good feeling for someone to care that much about what you’re building, what you’ve built.

Shensi Ding: Yeah. For sure. It’s very painful, like emotionally, but it’s good for people to give you that feedback.

 Kevin Rosenqvist: Yeah yeah, yeah, yeah. It’s a kind of a delicate balance there, I suppose. So. You mentioned unified APIs. This is a relatively new idea. Can you kind of explain how this approach is changing the game for businesses?

Shensi Ding: Yes. So previously whenever you would need to build integrations in your product, you’d have to hire engineers to build those integrations and you’d have to maintain them permanently. But to get to that point, there are still a lot of blockers. You need to get partnerships with different API providers that could take 6 to 12 months or even years. One of our partnerships has actually taken us four and a half years. And finally we got it now. And then once you get access to that API, you can start building and scoping the integrations or product. And design is often involved to try to figure out what that workflow looks like. You might build that integration. You might build it incorrectly. You might have to make modifications. You might also have to benchmark against what your existing integrations look like to make sure that the workflows look very similar in this new integration, too. And then you have to train sales enablement, and you also have to train your customer success team on how to build, how to use the integration and how to onboard customers. And then once the customer is onboard, there’s a lot of issues that could happen when it comes to comes to authentication, pagination or like edge cases and just a whole mess of things can happen. Instead of having to do all of that in house with merge, you don’t have to get the partnership necessarily with every single provider.

Shensi Ding: There are a few that do require it before you can get access, but many and most do not. So you don’t have to wait. You can just integrate once with one normalized API. You don’t have to understand the underlying systems, and you’re able to read and write data in the same exact format. So it just makes it way easier to offer those integrations out of the box in your product. In addition, you’re able to provide a consistent integration experience to all of your customers, regardless of what API provider that they’re using. Because if you start off with building your first integration, you build it so it perfectly matches that the capabilities of that first API. But then when you build the second integration, it’s not going to match exactly. And so you’re going to have different features that are available just purely based on what vendor your customer is using. And that gets way more complicated when you start adding integration. Eight 950 now your customer success and sales team has to memorize the functionality of every single integration, just because you because you didn’t really understand what the normalization or what the capabilities in the entire industry look like. And so instead of having to do that, merge is able to really just abstract it away into a common data model that is consistent and usually supported across the industry.

Shensi Ding: So you can provide that consistent experience. And if you want to add on top of it, because like you have a huge customer that insists on having one field, you can still do that through a unified API like merge is never going to be a blocker for you. You can still get access to that direct data. You can still make direct password requests, but it just makes it way, way, way easier. And that 99% is really handled through us. And integration, observability, which a lot of people don’t talk about. A lot of people talk about the initial, just initial build and just offering that integration. But when customer success has questions, when you need to add custom fields, we need to see what’s going on. Permissions that is usually done in like products like Datadog or Sentry or some kind of internal tool. And it’s very hard to get customer success access to that. And engineers have to usually take time away from whatever they’re working on to help look into these on call alerts. But with merge, we make it very easy. So customer Success has control over the integrations and has visibility. So unified APIs are really meant to handle the entire integrations lifecycle, not just the initial build, but the long term integration, maintenance and observability too.

 Kevin Rosenqvist: Okay. And let’s talk about what happens when integrations go wrong or when they’re not optimized. What’s the impact that you’ve seen on businesses that have poor, poor integrations?

Shensi Ding: Well, they end up having to take a lot of resources away from the roadmap that actually makes them unique. Like most companies, their product is not the integrations, it’s the data that they can build on. Or it’s like the experience that they can build on top, or the workflows that they can build on top of the data that comes back from those integrations. And any time that they take away from that, it’s not something that they’re going to be uniquely good at. And so for us, like what we’ve been able to do is allow them to just focus on their core product and also cut a lot of expenses. Like, you end up having to hire a lot of engineers to build and maintain those integrations, but instead, when you outsource that to a third party, you’re able to focus on your core product. What gives you high leverage and potentially also increase your revenue?

 Kevin Rosenqvist: What do you think companies kind of underestimate about the integration process. What are they? What do they not see? You know, from the cheap seats.

Shensi Ding: So either people think that like they just build it. It’s not going to have that much maintenance. It’s very easy. Like your engineers just throw up the integration and they’re good. That’s just not true. It’s never true. There’s always going to be a situation where the documentation isn’t correct. Response body comes back completely differently. You aren’t built for scale. You actually don’t know all the data points of what permissions errors could happen, and your integration just won’t reach maturity for a while. And I think a lot of people don’t think about that. On the other hand, I think a lot of people think that they’re a special snowflake and every customer needs a specific integration build. And that’s just not a great way to build a business. Like you want to have integrations be repeatable, and merge is a great way to do that.

 Kevin Rosenqvist: So with businesses becoming more interconnected, what role do you see APIs playing in the future of technology?

Shensi Ding: I think every company is probably going to have some form of a public API. Now. Everyone just expects the data to be flowing seamlessly between different applications. And if you’re dealing with a silo or only available via CSV download, you’re you’re just going to be falling behind. It’s just not possible for your data to be isolated, and especially with the prevalence of AI. And all these companies like using AI workflows and having different applications, you just you can’t fall behind. You need to have customer data in your product in order to provide better insights and workflows to them.

 Kevin Rosenqvist: Are there any emerging trends that that kind of excite or might even, you know, freak you out a little bit?

Shensi Ding: I think just the amount of data that every company needs is very scary. like a lot of these AI companies come to merge and they want every single category, every single integration, and they want every single piece of data that has ever happened in this company in the entire history of it existing. That is a lot of data. Obviously, you have to build a lot when it comes to our infrastructure and thinking capabilities and also figuring out how to price that appropriately. But like increasingly that is the demand of the industry. And so we’re just going to move in that direction.

 Kevin Rosenqvist: What about data security with more and more, you know, systems getting connected through APIs. What do you think the biggest risks they face are, and how can the industry address them?

 Shensi Ding: I think the future risk that will probably happen is people will authenticate their integrations and connect them, and they might leave the company and the and the head of security or CISO just has no idea that that integration exists, that to know that those credentials are there, to even know that the data is flowing somewhere else. And maybe you turn from a vendor and they have like live API keys, they’re consistently making API requests. I think that is something that people are going to start thinking about more in the future. We’re still actually in the early days when it comes to integrations, so I don’t think it’s a problem yet. And usually you require admin access in order to do that in order to get the credentials and share them. But I think long term in the future, especially as individual authentication becomes more prevalent, as more use cases come like start to rise, it’s going to be a little bit scary for a lot of security teams.

 Kevin Rosenqvist: Even even I think about my own, just like how I integrate, how I handle my own personal stuff and all the different things that talk to each other and make it work. Like, eventually I’m going to start forgetting what I’ve got going on here. So yeah, I can totally see what you’re saying. Somehow you got to be able to make sure that other people know that you have the checks and balances, I suppose, in your company. Startups often will struggle with scaling tech stacks as they grow. What should companies think about as far as integrations go to sort of future proof their systems?

Shensi Ding: Yeah, it’s always easy to be like, oh, you should build for scale. Like you should try to like make sure that this it’s like it’s isolated on its own. So where is it going to be easy for us to refactor. And the code isn’t messy and like and then we can continue to add on and layer on more and more integrations. It’s very easy to say, but when you have that customer that can make a big difference for your business and is saying that they’ll sign tomorrow, if you can start building and finish in two weeks, it’s going to be impossible to do. And I think a lot of startups either move too slowly, so when they die, don’t close their revenue, or they move too quickly and then end up having that tech debt later on and end up having to refactor a lot of their code base. So yeah, what’s the answer?

 Kevin Rosenqvist: Yeah.

 Shensi Ding: It’s a tough balance. Like it’s never going to be perfect. And there’s definitely parts of our code that like, of course, we could have spent like a few more weeks, like trying to figure out the perfect way to do it. But you need to survive. And that’s the most important thing. And sometimes you just have to move quickly and get that and eat the tech debt.

 Kevin Rosenqvist: What are some of the key lessons you’ve learned about solving complex problems, like software integrations, as you’ve gone through this merge journey of yours?

Shensi Ding: I mean, you have to do it on your own. And this is great Andy Grove thesis where, like as a manager, you have two responsibilities to make decisions or and also data collection. And I think the best way to collect data is to just do the job. So whether it’s like coding and getting to the code base or really understand what’s going on, or if there’s a leader and something’s not going right, like actually understanding, like what the what they’re doing day to day, what the team is doing day to day so you can understand what that looks like. It just gives you more confidence in making those decisions versus fully relying on anecdotes. So I think that’s one thing that, you know, a lot of people have been like, oh, like, it’s not a good use of your time to be doing this. Or like, oh, like you should be relying mostly on like, your managers. But I think the most important thing to do, you have to just do it so you understand what’s going on. And so yeah, I always make sure that’s not a boundary that I am going to be constrained from average.

 Kevin Rosenqvist: I and everything is, is creating all these opportunities for all these startups to come around and, and sometimes you think like people might not want to tackle those unsexy challenges, you know, the stuff that lives in the back end, the stuff that’s not as out there and stuff, and everybody wants to wear a turtleneck and be on stage and talk to a big group of people. What is your advice to someone who might be looking at a challenge to solve, like an integration type thing, but maybe, you know, it’s not as in the forefront as some other things might be.

Shensi Ding: I mean, that’s the best. Anything that is at the forefront. Everyone else is looking at that too. Anything that is extremely painful, unsexy, and like it’s probably going to not get that much attention is going to be a great idea. And I think anything that, like everyone else is doing is going to be harder to stand out. Everyone. There’s always going to be someone who’s smarter, harder working, better than you. And the less competition there is, especially with a really unsexy idea you can win.

 Kevin Rosenqvist: Yeah, I mean, it’s a good point. You know, I mean, because like I said, everybody wants to be that, that top, that top dog or whatever. And you still can be in other stuff that works in the back end. I mean, there’s a lot of functions that work in the back end that nobody really even knows about. And you think about even things like blockchain and all that. And it’s just there’s so many things that happen and, and it’s not exactly something that anybody knows, knows is even happening at the time.

Shensi Ding: Yeah. If you’re willing to get really in the weeds and get unique insights just based on pain tolerance, then yeah, there’s a lot of upside.

 Kevin Rosenqvist: There is. Was that a thought of yours as you guys went to, to build merge, or was it really just you saw this problem and you wanted to solve it? Or were you kind of like looking for something that was maybe a little off the beaten path?

Shensi Ding: I think to be honest, we just thought it was such an interesting idea and that no one had solved it. And the tailwinds are really going to help move us in that direction, in the direction where there is no future, where merch should not exist. That I think that was what really drew me, not necessarily just because it was on sexy, although that probably was a reason why it wasn’t as competitive when we first got started. When Merge’s success has obviously made it a little bit more sexy, which is why we started seeing more companies that are just copying the idea. But I think it was really just like passion for the idea and problem solving. And if you’re not really passionate about it, it’s hard to spend 10 to 20 years on it. And so I think especially if it is a unique insight and it is a unique idea that you have, it’s much easier to retain that passion versus when you’re copying someone because you’re just not actually genuinely thinking of something new. So it’s hard to have that motivation.

 Kevin Rosenqvist: Yeah. For sure. When you think about the technology as a whole, there’s a lot of people who are starting to get really nervous about stuff. They’re really nervous about AI, they’re really nervous about quantum computing for for good reason and all that. What makes you nervous and what do you think? Maybe that we’re, I don’t know, overemphasizing, maybe we should that people are nervous about that. They shouldn’t be.

Shensi Ding: I think overregulation on innovation is really tough because I think there’s overregulation and it’s very difficult for people to like, think of new ideas and make progress. I also think that when it becomes a moat for a lot of companies that are much larger because they inherently are probably helping craft the regulation. So it makes it so that they are the only ones that are able to be successful in the industry and these up and coming companies, it makes it much harder for like new competition for up and coming companies to be successful. And so I think, like, I know that a lot of people are very scared. And of course, like some regulation does make sense. But I think overregulation in new areas could be really dangerous just for new competition and just like a high like better like products for people to use and also increasing safety. So I think that’s like one of the things that I’m actually more scared of is just too much regulation just really can hurt innovation.

 Kevin Rosenqvist: Is there anything that you see everybody all scared about that you’re just kind of like, ah, pump the brakes. It’s not a big deal. Maybe you’re concerned about AGI. Is that something that freaks you out at all?

Shensi Ding: Uh, to be honest, I haven’t, like, gotten fear mongered into it yet. Obviously, like, I’ve seen, like, Black Mirror and all these, like, scary things that have happened. Um, and of course, like, it can. But I also think there’s just there are so many beneficial things that could be helpful that it would be a shame for us to suppress that before it gets there, even though it can, you know, obviously it could get out of control one day, but but yeah, there’s just so many potential benefits to it so far. It would feel like a shame if we suppressed it just because we were scared.

 Kevin Rosenqvist: Yeah, I agree with you.

 Shensi Ding: Scary. You know, like the human could have just not ever used fire because that’s very terrifying. But here we are today.

 Kevin Rosenqvist: That’s true. That’s a very good point. But. But Black Mirror does embed itself into my brain, and sometimes it is hard to forget it. All right. Well, sensei, thanks so much for joining me here and wish you the best with your merge. And, yeah, thanks for sitting down.