Into the Forge Podcast Season 3, Episode 5: Dave Gustafson, Omar Khan, Garrett Larsson of Rhombus
In each of our podcasts, we ask top hardware entrepreneurs the same 10 questions to better understand the challenges and best practices in starting a hardware company. In Season 3 Episode 5, Lemnos’s Eric Klein speaks with Dave Gustafson, Co-founder and VP of Hardware at Rhombus Systems, Omar Khan, co-founder and CTO, and Garrett Larsson, Co-founder and CEO of Rhombus Systems, a Lemnos portfolio company. Rhombus leverages AI and computer vision to offer video security systems unlike anything else.
- What compelled you to start your hardware company?
Omar: We were itching to start a company right after our company got acquired by Sophos. It was called Mojave. We were bringing basically firewall to the cloud. Once that got done, in the process we saw that people wanted to bring in more hardware to the cloud. That’s where we saw a few opportunities. We learned there’s a missing opportunity where the camera market is growing a lot in the consumer space, and nothing has been done in the enterprise space. That’s where it started all happening. Garrett and I got together, started discussing out things, and that’s when we saw that we can do a hardware company. Then we found Dave with the hardware expertise to help us do that.
- Had you worked on hardware projects before this startup?
Dave: I’ve been doing hardware my entire career. I was at Frog Design and the nice thing about that is to bring a design sensibility to an engineering task. After that I was making drones for the commercial sector for mining, construction, agriculture, those kinds of things. That’s definitely a very intense hardware situation, which has to do with safety and reliability, but also imaging. So, taking some of the specialized imaging that happened on drones and moving it to the different kind of specialized imaging that we’re using for security cameras now, has been an interesting process.
Omar: My research background was all embedded systems. But my actual experience after college was all security-related, ranging from web application security to mobile security to cloud firewall. So, it was just a gel of all those things that I think worked out pretty well. Even though I hadn’t worked on a hardware company before, it was intuitive enough to work on embedded systems going forward.
Garrett: I probably echo Omar. Barely touching hardware, maybe doing a little bit of embedded systems. When we were first starting up, just the idea of having hardware was daunting. What I’d say for first time founders out there, even though it was daunting, coming from a technical background, doing some of the embedded stuff before, some of the firmware stuff, it wasn’t as daunting as it seemed. And it probably helped also to get a co-founder right away that had hardware experience to pave the path a little bit.
- How did you decide what would be your first product?
Garrett: The first thing was we identified a space that we wanted to be in. Omar and I were doing a lot of enterprise stuff before. That was where we were comfortable. We’re not consumer guys. We’re not trendy like that. So we didn’t know what the next hot hardware item would be in that area, but we had this thesis in enterprise that there was a lot of archaic hardware that needed to be brought to the cloud. And then we got to this physical security space, and we felt like that kind of stuff should start happening on the enterprise. Then we started looking at the different things that go into physical security. We thought cameras would be the best entry point. From there, we felt customer interviews were really important. So, then we started talking to IT people, which we thought would be our target customer. “Hey what do you think of this system?” We validated with them that that was a good entry point.
- How did you decide who would be your mentors?
Omar: We knew that we needed an advisor on the hardware side. So that was no-brainer.
Garrett: Omar would keep sending me links on LinkedIn, “What do you think of this person? What do you think of this person?”
Omar: Exactly. So once we’d narrowed down who we wanted as an advisor, it was hard to get to them. It was again through networking we met someone. I would say we had good success at least having the hardware advisor, helping us make those intros that we were looking for with ODMs, JBMs, or these tech manufacturers. I would recommend where a company sees they lack …
Garrett: Core competency.
Omar: Exactly. So that’s where they should look for good particular advisors and mentors.
Garrett: At our previous startup, we probably had 10 or 12 advisors, and they all give you random advice. Usually as a startup you’re giving them all a little bit of stock, and they just don’t do much, so just focus on a couple that you really think bring value.
- What have you gained from working with Lemnos?
Dave: I think this is another case where having hardware expertise was extremely important to us. We realized that even though we’re trying to keep the hardware side of our operation on the lean side, that’s still the area where you can make mistakes that are costly or keep you from having a product on the market, or cause recalls or inventory shortage or overage—all those kinds of things. The best thing you can do is have somebody who’s been down all those roads and can give advice on it. So it was a de-risking thing. You guys completely embrace it. You live and breathe it, and that bit of expertise was really important to us.
Garrett: By the time we got to you guys we had first customers, and we wanted a partner that could help us scale our hardware ambitions, and also scale the business as we looked to Series A.
- What was the road to your first round of financing?
Dave: I kind of go back to the hardware aspect of it, where I’m trying to make investors and everybody comfortable with the notion that we’re not trying to do something that is hardware risk. We’re not trying to do anything that is fundamentally different than any hardware that’s out there, we’re not pushing any envelopes, because it costs a lot of money to push hardware envelopes. Instead my perspective is always, “Can this problem be solved with software?” If so, let’s do that, and we only get to making our own hardware once we’ve exhausted the other possibilities. That’s my way of keeping it lean, keeping everybody as comfortable with the risk portion that’s being taken on by hardware in the company.
Garrett: Then in terms of how that all formulated into pitches, we’d done the pitches a couple of times, so I guess we had familiarity with it. I would say even with familiarity, you’re learning every time. No two investors are the same; you can’t ever pitch the same. That’s a good place to also look for mentors—people that are in the industry or have pitched a lot.
- What has been the most surprising thing about the manufacturing process?
Dave: As far as scaling, there are places along the scale where certain processes need to change. When you start off, I’m building them all in my garage. When something goes wrong, they’re coming back to me and I’m fixing them. It’s all me. So I need to figure out when I put the task off to another party. So obviously we’ve got our manufacturer who’s shipping the product to us, we’ve got to forecast accurately against our sales, we’ve got to engage a third party logistics provider. They’re going to be more robust return and replacement and diagnosis of problem processes that need to be put in place. So that’s going to be continuing.
Garrett: Well, at first we thought it would be very easy to get an ODM partner. That took us longer than expected. And as the non hardware person in there, I was actually surprised at the speed we were able to go from “Yes, we’re doing this,” to writing code, to getting hardware, to getting the molds done or the outside casing. It came together a lot faster than I would have expected. If you had asked me right before we started the company if we’d have been shipping beta units out to customers in six months, I think I would have probably guessed one year.
- What have you learned through customer engagement?
Garrett: Let’s try to talk to as many potential customers as we think we can, and we did that very early on. Just to make sure that, yes this is something that we could actually sell. Not all those first customers ended up buying. You get random feedback from random people. And then, iterating as you go. So, we talked to them before we had a product. Then, the first beta product we had it’s like, “What do you think of this?” And then the next alpha, or the next betas, “What do you think of this?” Just keep trying to iterate on that. There’s always this healthy tension between what customers are asking for and what’s best for the company. Sometimes what customers are asking for is best for the company, and you should attack that. And then sometimes you have got to be able to say “No.”
- What’s the hardest part of being a hardware startup founder?
Dave: I thought it was surprisingly difficult to find a manufacturer who was willing to engage with us in the way that we wanted to engage. It really came down to networking, advisors, personal relationships, and finding a company that was willing to take a bet, because our order sizes compared to their manufacturing capability is quite small, so the fact that they’re investing any of their resources and doing the kind of customization that we require is them really betting on us as a company. We imagined the OEMs would be open to whatever, as long as you’re paying the price, but not so much. Now that we’ve found it, it’s a good relationship.
Omar: For me, the hardest problem was making sure that once the hardware, the firmware is released from our nest. We always have a backup if things go wrong the way that we didn’t perceive. We can fix them without incurring a big cost. To everyone who is doing it first time, things don’t go as expected. They didn’t go, for us, the first time, for our first batch of hardware, and fortunately enough we had a lot of sleepless nights before that. We made the conscious decision to have three, four different scenarios to get onto the camera and fix any possible problem that would arise.
Garrett: I think a lot of technical founders can get into the trap of thinking, “I will build the greatest product and customers will come.” And I think we get reminded and humbled every day. There’s a big sales and marketing arm that you always have to build. You have to think about sales, and you have to think about sales early. Sales would be, for the consumer, just as important to your organization, at least for enterprise, as your product. I think sometimes, living in Silicon Valley, you sometimes under-appreciate that.
Omar: I have never valued more what sales and marketing were doing until I became a founder and saw first hand. I would literally say it’s 60/40, like 60 percent is sales and marketing, 40 percent as tech and product. So you can have the best product in the end-price space.
- How do you find a much-needed work-life balance?
Omar: I’ll be honest, it’s pretty hard for me to have a work/life balance in these early days.
Dave: I’ve got a couple of small kids, so chances are if I’m not working, I’m parenting, which is great. I think the nice thing about that is that little kids, they force you to adapt to all kinds of different situations. Your brain is in different modes all the time, depending on what they’re doing and what they need. The nice thing is that there’s usually some little other background part of your brain that can still be working on some kind of work problem, and maybe even receive helpful outside triggers from what’s going on with parenting to look at things in different ways.
Garrett: I would agree that the family helps bring some work/life balance. We are coaching Omar on all this. For myself, I need to hit the gym, or run, or something. It’s a natural time to just take a break and think, and even Omar, I know you will say this, “Oh, I put the problem aside for an hour, and I came up with the solution!”