How we built OfficeSpace Assets with AI (and heart): A Q&A with our engineering team

Meet Ly Dang and J.F. Turcot, the husband-and-wife engineering team behind OfficeSpace Assets—our powerful new enterprise asset management platform built to help facility and IT teams manage workplace assets with unprecedented ease and efficiency.

Together, led by CTO Kiley Reynolds, they turned a moonshot into reality—delivering a world-class solution in just 74 days. With the help of AI, deep product intuition, and a whole lot of heart, the team brought OfficeSpace Assets to life at a record pace.

OfficeSpace CMO Heather Larrabee sat down with them to learn how it all came together, what it means for our customers, and how it’s shaping the future of our platform.

Q: Ly and J.F., you have a unique story about how you met and how OfficeSpace brought you together. As partners at work and at home, tell us a little bit about how you met and how we got here today.

Ly: We actually met at OfficeSpace. Our love grew from mutual respect as engineers; we loved how each of us worked. Our marriage even began with AI! We were just going to have Huey, our principal architect, sign our marriage certificate two years ago, but last minute, we decided on a small ceremony because everyone was here. Huey quickly put together a ChatGPT speech an hour before, and we had an intimate ceremony with our amazing coworkers and parents. That’s how we began at OfficeSpace!

Q: J.F, before this fated day, how did you actually make your way to Ly?

J.F: Our love really began after I drove down for two days from Canada. I just wanted to come down and see my team in Atlanta. I didn’t tell anyone, just got in the car. On my way there, after driving for two days and 15 minutes away from the hotel, I hit a giant pothole. I had to call Ly to save me. She actually knew I was coming down, but she came to save me and drove me back to the hotel. That’s how it all began, with me getting into a small car accident.

Q: Kiley, you brought Ly and J.F. together and saw something in their work styles. What did you see, and why did you have them working together?

Kiley: Well first of all, I was very surprised when J.F. showed up in the office! But what I saw in both Ly and J.F. is something truly hard to find in anyone: the caring, ownership, and pride they take in their profession. It’s not just a job; it’s their profession, and they genuinely care about it. When I’m looking for developers to drive a project from beginning to end, who will truly put in the time and effort, I couldn’t think of any two people better than them. They’ve shown that commitment since day one.

Q: Let’s talk about the call to arms that led to this project. Our CEO attended a summit in February and returned inspired by the power of agentic AI. How did this spark the rapid development of OfficeSpace Assets?

Kiley: It’s actually a pretty fun story that began even before that. We’ve been pushing hard into AI for about a year and a half, constantly exploring how we can improve. Fast forward to Super Bowl weekend in February: our CEO, Erin, started texting and calling me, excited about agentic AI’s potential for coding. I then spoke with Andres, our CPO, and Huey, our principal architect, and Mitra, our Director of Engineering. Over about 48 hours, we were incredibly excited, looking at tools. We found Cursor and decided to just build something in 24 hours to see what was possible.

By the end of that weekend, we had three different Proofs of Concept (POCs), and asset management was one of them. Our initial high-level estimates suggested asset management would take a couple of years and two teams to build, costing millions. But Huey took our high-level requirements and built a fully stood-up POC over that weekend. Our eyes lit up. We realized we could build this, and build it very quickly. We set an audacious goal: deliver it in three months, something no one thought was possible. I then chose Ly and J.F, knowing their expertise and startup history made them the perfect combination to tackle this game-changing opportunity. They spent 74 days hands-on, didn’t cut any scope, and delivered what we thought was impossible.

Q: Ly, you started your build on April 11th, and 74 days later, a world-class enterprise asset management solution was delivered. You’ve spoken about how much AI coding assistants have evolved. How did you begin to build this momentum, and what were some key learnings?

Ly: When all of this started, I never believed AI would get to where it is today. We began with GitHub Copilot, and in its early days, it barely functioned for production-ready environments, often hallucinating. But fast forward two years, and you can now build entire applications with it. It’s truly mind-blowing; I don’t even know how to describe what we accomplished.

Q: We know AI coding assistants still require immense subject matter expertise. How did you ensure that what you created for users actually helps them accomplish their jobs, avoiding scope creep and maintenance issues?

J.F: We never do “vibe coding” here; we do AI engineering. We take the time to build proper plans and ensure everything is right before writing a single line of code. We go back to basics, focusing on proper engineering, not just building a quick to-do list like you might see on YouTube. We built a real product with real architecture that has been thoroughly thought through.

AI is a lot like the old Simpsons episode where Homer is asked to design a car and he just goes absolutely wild with it. He goes completely overboard, making it look ridiculous. AI does the exact same thing. It’s amazing; it can suggest something you hadn’t thought of, saving you months. But it also suggests hundreds of other things we don’t need right now. My job is to ensure we only pick the good ideas and avoid overbuilding. Everything we build that isn’t needed still requires maintenance and poses a security risk because it’s there but untouched. It also impacts reliability, as something unnecessary might crash the system. So, my job as an engineer is to ensure everything is well-architected and stays focused on the goal.

Q: OfficeSpace transformed its Early Access Program due to our rapid building pace, allowing real-time feedback from clients like Moderna and WWE. What was it like to get this rapid feedback and immediately roll it into what you were building?

Ly: It was incredible. I waited my whole career to do this. I left IBM to be closer to clients because I wanted to hear what they said about what I build. To hear them on a weekly basis tell me they don’t like this, or they’d like to add more of that, and then actually be able to do it and show it to them the very next week, is unheard of. In the past, we’d build something, maybe show it five months later, finally get feedback, and then deliver what they asked for a month or two after that. So to get that feedback weekly, and to hear how it’s solving real-world problems for clients like Luke from WWE, is absolutely incredible.

Q: What does this rapid, client-driven innovation mean for client experience, impact, and ROI? How will this approach change the game for businesses?

Kiley: It changes everything. Being able to listen to our clients in the moment while we’re building something, and then deliver massive features and changes within a week, is unprecedented. Typically, it takes us years to build a product like this, and then we’d get feedback in beta programs, with feature requests taking months to implement, if they happened at all. This new approach means we can incorporate valid client feedback and add large, valuable features to our products within a week or two.

Product development is now on another level. We’re not just delivering the current product; we’re thinking about the product of the future. Our architecture is built to enable us to build new things faster and integrate AI. The script has flipped: it’s no longer just sales pushing engineering for features. Now, everyone else has to keep up with how fast we can build. If you’re not using these tools, you’re doing something wrong, and clients will go to those who can provide value quickly. Even if we don’t have something today, we can deliver it in the very near future, which most companies can’t say.

Q: J.F, you’ve used a chess metaphor to describe the complexity and intensity of this process, moving with unbelievable speed while predicting future needs. How do you approach such a design challenge?

J.F: We plan a lot. Even with AI, it’s a conversation, not just me writing. We take our time to ensure design documents are properly written—something we didn’t have time for in the past. The AI is incredibly intelligent, sometimes too much so, suggesting hundreds of things we don’t need. My job is to pick only the good ideas and avoid overbuilding, because anything unnecessary still requires maintenance, poses a security risk, and impacts reliability. We built this new product as the backbone for everything at OfficeSpace, prioritizing reliability and security.

I’ve played with AI before, but Kiley really pushed us to use it. Many of us were skeptical, thinking AI code was bad. But he was right. I couldn’t have done what we did in two months without it; the quality of the code is truly impressive if you keep it on track and approach it as a proper engineer, not a “vibe coder.”

Q: Ly, you pushed through tremendous stress and deadlines as a working parent with three young kids, working around the clock even with AI’s help. How did you find the personal resilience to do all of that?

Ly: I’m still reflecting on how we did it; we just powered through. This project became our baby, and we wanted it to succeed. We took every bug personally and were determined to meet our goal. We’re both very competitive and didn’t take failure as an answer. I feel bad that I didn’t spend more time with the kids, but I believe I built something they will be so proud of in ten years when I show them all of this.

Q: J.F, what about you? How did you survive this 74-day gauntlet?

J.F: It was fun. That’s how I survived. It was a lot of work, yes, but it was rewarding every time we pushed something new. Earlier today, JD, our security guy, showed me a graph of everything we built, and I looked at it like, “no way did I build that in 74 days.” It’s really impressive. The fun parts of engineering, like building stuff, now dominate. Finding bugs, the not-so-fun part, doesn’t really happen anymore because AI finds them so much faster. It’s like building with Lego blocks without ever losing a piece under the couch—just building something new and seeing the results.

Kiley: There are two things I want to emphasize. First, Ly mentioned they are both competitive, which was a key factor when I chose them. Their competitiveness drives them to continuously push forward, which is amazing for a project like this. Second, what J.F. said about creativity is crucial. All of us in engineering got into this field to be creative. In many companies, building a single feature is a long, 6-9 month slog, leading to burnout and boredom, which actually slows things down. AI is allowing us to rediscover that creativity, passion, and energy that many of us older engineers, like J.F. and myself, have felt was missing for a long time. It’s like a rebirth, a re-envisioning of how we do things, and it truly makes it fun again.
Ready to transform your asset management from a challenge into a strategic advantage? Discover how OfficeSpace Assets can streamline your operations, reduce risk, and provide the peace of mind you need. Try out the product today and see the future of asset management in action.