Tackling the unsolved problem of BPO management: A Q&A with Assembled’s most tenured engineer

Assembled just hit a significant milestone: 100 employees. In honor of the company’s journey from three Stripe alums in 2018 to triple digits in headcount six years later, we’re spotlighting early hires and the unique paths they’ve created for themselves at Assembled. You can find all their stories here.

Taylor Milliman is an Engineering Manager who led the team that built Assembled Vendor Management. In this interview, he shares what it was like working at Assembled when it was still in its seed stage, how the company has managed to maintain its scrappiness as it’s matured, and why a little bit of paranoia is actually a good thing when building a new product. His answers have been edited for clarity and concision.

Q. Taylor, how did you end up at Assembled?

A. It was the beginning of the pandemic, and I always knew I wanted to join a startup and leave my current company. I asked around about which startups I should be talking to, and my friend Jordan recommended Assembled. At the time, it was smaller than what I was looking for — still seed stage with eight people.

My friend Edwin then introduced me to to John (Wang, co-founder and CTO) and Jen (Ong Vaughan, first business hire and former COO) who he had worked with at Stripe, and spoke highly of them. After my phone screen and meeting everyone on the team, it just felt perfect. I was living less than a quarter mile from the office, which was an apartment at the time. I really enjoyed meeting everyone, and during the pandemic, my risk tolerance was higher, so I just went for it.

Q. What do you feel has been your most significant contribution to Assembled?

A. I see myself as a product person, so I'm most proud of the collection of product improvements I've been a part of. If I go through Assembled's pages, I've probably touched at least 50% of them in some material way. I'm really proud of the breadth of my contributions and how much I've shaped the product in the few years I've been here.

Q. What's the most striking difference between Assembled's product when you first started and now?

A. The most striking changes aren't the ones I've made. I'm really excited about the work Luke (Andrews, Head of UX) has done on the look and feel of Assembled. When I joined, we used an out-of-the-box bootstrap theme that looked good at the time, but now we realize it didn't. The evolution in the look and feel has been really cool to see, and it’s also become a major differentiator over time.

Q. In those early days, what were some of the most challenging technical problems you worked on?

A. One of the first was Google Calendar. It doesn't sound that challenging, but setting up a two-way sync turned into a distributed calendar problem where we didn't know whether the source of truth was Assembled or Google Calendar. There were so many edge cases. It was one of the first major projects that sounded easy but turned out to be super hard to get right.

Another challenge, which is honestly still ongoing, has been building a really good ticketing API. I built one of the first versions for Stripe and a few other customers. It's a tough API to get right because it needs to be compatible with different CRMs. One of Assembled's major value props is ensuring accurate reporting, so designing an API as good as our native Zendesk or Salesforce integration is pretty tough.

Q. How do those challenges compare to the technical problems you're working on today?

A.  Today, the problems are more about scale than zero to one. Performance is a huge focus now, which wasn't as big of a deal when we had smaller customers and less data. For example, with vendor management, performance is critical because it's used by our largest customers. We also deal with complex migrations and minimizing downtime, which is more challenging now that we have major enterprise clients who expect high reliability.

Q. How did the need for a vendor management product become apparent?

A.  It goes back to before I joined the company. One of Assembled's first customers was Stripe, and part of how we convinced them to sign with Assembled was by promising to integrate with their BPOs. This was back in 2019, and at the time, we created these really scrappy, non-scalable integrations to pull in their schedules. It was barely a product — super scrappy. Some people at Stripe didn't even know these integrations existed; they just worked in the background.

We've had the idea for a long time. Stripe was one of the first customers pushing us in this direction, and for a while, almost our entire product was kind of unusable for them except for the vendor management features. By 2022, we realized we had three options: let Stripe churn, build a bunch of custom features for them, or build a real product that we could scale for other customers.

This decision isn't always straightforward — sometimes building a new product isn't the best move. But in this case, vendor management is really just workforce management at a higher level. Workforce management involves managing people and teams, while vendor management is about managing vendors who manage teams who manage people. It aligned well with our company vision and our goal to go upmarket, as nearly all customers with over a couple hundred agents start outsourcing. So, we decided to invest in it.

Q. How did you end up leading that project?

A. You know, that's a good question. I remember that John asked me to lead it as we were wrapping up another project. I initially thought it was a bad idea and was super skeptical. Up to that point, I had mostly worked on scheduling and redesigning some of our pages, so I didn't know much about vendor management. I didn't have much background in it.

It was probably a combination of other people being tied up in other areas and John's confidence in my ability. My skill set is very startup-focused — I work very full-stack and get involved with the design of things, not just the technical contributions. So, my background suited the project well.

Eventually, after hearing more about the opportunity and talking to Stripe about their problems, I started to get excited about it. It wasn't really a question of whether I wanted to do it — it was more of John telling me to do it. Once I understood more, and especially after talking to real customers and hearing about their problems, I realized there was a huge opportunity to do something impactful.

Stripe wasn't demanding that we build this; they were more like, "This is a huge problem for us, and we think you can help." That motivated me even more.

Q. Vendor management came together quickly. What did you do to get there?

A. The first version was probably built in about three or four weeks. Our basic idea was to clone a spreadsheet that Stripe was using but add some unique advantages to make it better than a regular spreadsheet. We demoed it to Stripe around September 2022. The initial iteration was quick, but we didn't start getting serious traction with Stripe until about three months later.

In hindsight, we could have moved faster, but at the time, it was a mix of strategy and experimentation. We had ideas about what we could improve, but a lot of it was trying things out and figuring it out as we went. It was a combination of strategic planning and on-the-fly problem-solving.

Q. Is it accurate to say vendor management is a first-of-its-kind product?

A.  Yes and no. I think the BPO planner is more of a first-of-its-kind thing. But the overall problem of vendor management has seen some attempts by other companies. For example, NICE has tried to build something in this space and has gained a bit of traction.

However, in terms of widespread adoption, it remains largely an unsolved problem. So, in that sense, our solution is relatively new. The same goes for integrating with BPOs — no one had really talked about it the way we did before our launch. To give credit to other competitors, they have tried to build related solutions, but I believe our approach and the timing were unique.

Q. How do you find inspiration to tackle unsolved problems? And how do you motivate your team to do the same?

A. I think what personally excites me is strategy. Figuring out why Assembled can do this better than anyone else is really motivating. A lot of it came down to the fact that we already had a lot of the data about vendors. What we didn't have, we could easily pull in and display in our existing UI.

Having this data and adding a few new pieces, like schedule and adherence data, gave us an edge. These insights are difficult to replicate in a spreadsheet because managing millions of events in a spreadsheet is a huge pain, and calculating adherence would take forever.

The strategy got me excited, and my philosophy is that the most exciting thing to people is progress. If you make fast progress, it builds on itself and becomes more fun. We started with the idea of building an MVP in three weeks. Keeping that momentum going makes progress fun, which is a big part of the startup experience. It builds excitement and makes you want to build more, and soon, people start coming up with their own ideas as well.

Q. Were there ever any moments where you were like, "Oh no, I don't think this is gonna work"?

A. Building a new product is pretty much just feeling like that almost every day. There's a cliché that "only the paranoid survive," and I feel that's definitely true with something like this. So, basically, every single day, I was worried about reasons why no one would use it or why no one would want to buy it, right up until launch.

The insight we eventually had was that at some point, you can't think your way out of the problem of whether someone's going to use it or not. You have to test it and see if people are willing to buy or use it. The market will decide.

I would get into arguments with the team all the time about what direction we should go, why certain ideas weren't valuable, or why they were. Eventually, we realized that if you're debating for too long, you probably just need to go build something. Especially for the first three months, but pretty much right up until launch, we were constantly debating that question.

Q. Any specific moments when you were super proud of customer feedback?

A. Yeah, definitely. The obvious one was our trip to Ireland to visit Stripe. They had a support team summit that week, and we hadn't met any of these people in person before. We worked super hard to get a good demo together, literally merging code during the demo to ensure everything worked and included all the features we were aiming for.

I remember being in a room with about 20 people from the Stripe team, including the head of WFM, and demoing what we had built. Seeing them get excited was incredibly rewarding. They also had an internal engineering team working on some vendor management stuff, and they told us we were much further along than their internal team, which was super exciting to hear.

We had all spent so much time on it in the previous months that I could demo it without thinking about what to say because we had been immersed in it every single day. That was a huge moment. Additionally, forming personal connections after showing the product was cool. Next thing you know, you're in Ireland at a bar, drinking a Guinness with the head of support at Stripe. That whole trip was a significant turning point for us.

Q. Prior to joining Assembled, had you been in a leadership or mentorship role?

A. No.

Q. Now that you've developed into that role at Assembled, what's the biggest thing you've learned about yourself?

A. I think the biggest thing I've learned is that I still get a lot of joy out of building new products with fun people. Seeing others experience what I did in my first two years here — having the opportunity to build, build, build, and put things out into the world — is incredibly rewarding. It's super cool to push people who report to me to experience the same growth and responsibility I had when I first joined.

When I first joined Assembled, I looked up to everyone around me — they were like the coolest people I'd ever met. They pushed me in ways I hadn't experienced before and gave me a lot of responsibility. Now, getting to pass that on and give a lot of responsibility to others in a similar spot to where I was two years ago is really fun.

Q. What's been the most striking cultural transition from a team of less than 10 to a team of 100?

A. It feels like a real company now, yet we still maintain a lot of the scrappiness and early-stage mindset. One of the best things I've seen with all the work John's doing on new products is that it really motivates the rest of the engineering team to keep that mindset. You can still ship quickly, come up with your own ideas, and put them into the product.

On the other hand, performance and stability have become much more critical. We have to take our uptime a lot more seriously now. So, it's about maturing as an organization, especially as an engineering organization. We need to keep the fun, quirky elements from the early days while recognizing that we have a job to do for our customers, which includes not taking down the site very often.

Q. What advice would you give to someone who wants to join Assembled as an engineer?

A. I think it has to be fun. Think about it as a journey you're going on, and you'll get a lot more out of the experience. If you naturally put a lot of effort into your work, you'll build great friendships, take on more responsibility, and grow. Follow what excites you and what you're interested in.

Also, take advantage of the healthy amount of chaos here. You can do whatever you want to some extent. You can either take advantage of that in a positive way or wait around for someone to tell you what to do. In my experience, growth and relationships come from taking initiative. Having fun with it is key.

Remember, ideally, you're here for a long time. If it's not fun, you'll probably leave after a couple of years. Give yourself a break sometimes and explore your own ideas.

Assembled is still growing! Check out our open roles.