Fiona Miller needed proof she was doing important and valuable things for electronics engineering, and did she ever get it.
Fiona Miller is a legend in the world of electronics engineering. After selling her last company a few years ago, she pivoted to run a venture fund that invests in biotech companies focused on age reversal.
Miller’s first introduction to engineering was in high school in 1978. Computers existed but were uncommon, and nobody knew what they were. “One nice teacher brought a very primitive little computer into our math class; the computer read paper tape and had a single-line LCD that could display seven characters. This teacher let us earn class credit by playing with this fascinating computer instead of doing math, and it pulled me right in,” she says.
Miller went to Columbia University’s Barnard College to major in philosophy yet took one computer-science course as a standalone elective. The passion for computers came back, and she found she could play for hours programming the primitive computers.
“In fact, programming is a lot of fun and relatively captivating,” she says. “But I began wanting to know more about the electronics behind these magical capabilities — commands and programmed responses to make people laugh and so on. So, I transferred to the engineering school at Columbia to study electrical engineering. Coming from a liberal arts school, I wasn’t really prepared for engineering coursework … but curiosity drove me.”
Recently, Design World caught up with Fiona Miller about her experiences at college, her work at automated test-equipment supplier Teradyne after school, and what drove her to her various successes after that. Here’s what she had to say.
Was your university a supportive environment?
At school, there were only two of us women in our graduating engineering class of about 300 engineers. So, we felt a little odd … but the other woman became valedictorian of our class. So, it was clear we could engineer. We just got used to being in classes full of guys, and nobody was particularly encouraging or discouraging — we were simply treated as adults. However, in my last year, I wanted a hands-on project to fully grasp the abstract things that weren’t quite sinking in for me. So, one of my professors let me pursue that as my coursework for the last semester. Well, the guy who ran the electronics lab was a nice old guy and very supportive. Because everything was very primitive at that time, I had to etch my own circuit board and expose it in a darkroom. He got me all the supplies and showed me how to do it all. We ended up building a little autonomous vehicle and had a ton of fun.
In fact, while in school, I also worked on programming to be used on Wall Street. At the time, no computerized systems were used to run the stock market. It’s very strange to think about now. A Columbia graduate working on a brokerage project bought a PDP 11 computer and put up a help-wanted advertisement looking for programmers. So, I applied and got a part-time job working for him for two or three years. I vividly remember the first project was to write a program to analyze option yields for options trading. My boss ran around with this printout that I’d produced saying, “Now we can do this report every night instead of intermittently doing it by hand.” We were part of the stock market’s initial early use of automation and computers.
But software was too easy for me, and I really wanted to do hardware. In 1983, when I got recruited by Teradyne, I moved to Boston … and that’s where I still am.
After a year and a half at Teradyne, I joined a startup called Chipcom (now 3com) as the seventh employee. Working late nights with founders, I was there to see the initial company setup and laboratory buildout with new equipment.
What did you learn at this small operation?
Ethernet is all integrated into computers now, but back then, Ethernet hardware consisted of cables and external boxes … and was mainly for local area networks and cable TV. We were building systems to carry 10 Mbps — a big step up from telephone-line modems capable of something like 128,000 baud and bits per second. At the time, I thought why would anyone need so much data speed? But I really liked being in a small company where I was very close to creating something new — and not just working on projects that were a small piece of a big system.
I always had the bug to start my own company. Some engineers enjoy working on the bench because getting things to work is kind of like play — but for pay. In fact, I know a lot of engineers who liked that. But to me, it was always important to do something that matters … my logic was that if I don’t make a lot of money building whatever I’m building, then I’m not doing anything that matters. If I get if people to pay me a lot of money for what I do, that means I’m doing something important and valuable. At larger companies, I did whatever marketing thought we needed to build, and it was always marketing people acting as decision-makers who decided what I would work on. But sometimes, a project I was putting my heart and soul into would get cancelled, and it was clear the project didn’t matter … and all those late nights working didn’t matter, either.
So, it finally dawned on me: The only way I’m going to work on something that matters is if I decide what it is. Little did I know that even if I myself decide what to develop, it may or may not be a thing that people want to buy. Over time it became clear: Choosing one’s own project means if a project commercially fails, you have no one else to blame but yourself.
After Chipcom and another job, you established StarTek to supply network and cable analyzers. Eventually, local area network supplier 3com (acquiring Boston-area companies) bought StarTek, but you were spun out with a new company as VP of engineering. What happened next?
Because there was no CEO at the time, I basically ran Scope Communications. In 1998, that company sold to Hewlett Packard, which at the time had a test-equipment business.
HP was very supportive of women, and the so-called HP Way imparted business agility while supporting engineering innovation. In 1999, however, Hewlett Packard and Agilent split, and my division became part of Agilent.
There were a lot of changes and not for the better. In contrast with HP, where people had a lot of freedom and engineers were encouraged to take initiatives within their own small self-managed divisions, Agilent embraced a more conventional and centralized organizational model. It just wasn’t the same. Then, in the 2000 tech bubble burst, there were huge layoffs. I decided to volunteer to take a severance package offering eight months of pay. In that way, I had eight months of freedom to create a business plan and get venture funding for my second startup — Azimuth Systems — sold to Anritsu in 2016. Next, with my own money (and no venture capital) I started octoScope essentially in my basement as a consultancy. That sold to Spirent in 2021.
Of which of your businesses or technologies are you most proud?
I’m most proud of octoScope because it was all mine and, therefore the most challenging … and yet I was already a mature professional. I’d worked at startups. I’d managed engineering groups. I’d raised venture capital and knew I didn’t want to go that route with this business. I had to build a team, and we were a good team; octoScope was a tremendous experience. At first, it helped that I was already well known in the computer-networking space. Soon, I had more business than I could handle. So, I got two super-smart engineers to join me as cofounders.
Together, the three of us got the consulting business to a million dollars. Then in a move away from the consulting work, we developed a product. Of course, then we had to figure out how to put this product into production and sell it. We also had to build our team because sales were quickly growing. So, I pretty much went to business school right there — a hands-on school that forced me to read a lot of business books and figure out ways to scale the business up to $20M. Then we sold to Spirent … and then I didn’t have to work anymore.
Why did you choose to work in the field of WiFi?
Well, WiFi is related to Ethernet and LANs with which I was already involved. WiFi was billed by IEEE as wireless Ethernet before it was recognized it would need to evolve away from Ethernet to work in a medium that’s much more challenging than cable. All the cellular networks in different markets were subject to different standard bodies … though ultimately, they did end up converging. WiMAX once served as the main intermediate between WiFi and cellular communications; then came LTE, 4G, and 5G. A lot of the technology for WiFi and cellular is now quite similar, and Ethernet makes many systems about as seamless as it gets. But back when I went into WiFi, it was not seamless; WiFi was managed by different standards bodies, and it was a sea of acronyms and technologies. Military was always 10 years behind everything else, and that’s a market unto itself. So, yes — I mostly stuck with WiFi until the networks started converging; then, I started drifting to more work in cellular networks to eventually come to do a little bit of everything.
What was most challenging — the actual engineering, managing a team, or raising capital?
Everything about raising capital for my second startup, Azimuth, was challenging. I worked with a couple of venture funds here in Boston. We had to gather information for them and learn to size up the market, competition, and customer base — all that. Even creating a business plan was a new challenge because, as an engineer, I’d never done this before. So, it was an exciting learning experience. After I raised the money, a CEO was put in place, and I lost a lot of control. Even so, I stuck it out for five years. When I left, I formed a consulting business because I felt like I needed my own space and to determine my own course. Such freedom just went away after venture funding.
In contrast, early on at octoScope, we faced a serious technical challenge. We created a small test system in which we emulated real-life environments for residential and other wireless communications. People used to run around their houses and test their WiFi at different points. Well, our job was to test WiFi setups in a small enclosure … and get that enclosure to behave like a big house. The multi-factor physics of a house involves places where there are walls and other reflection points (such as those from furniture) and varied travel distances based on how big the house is. We couldn’t just test WiFi devices in clean test-enclosure environments and expect to catch all potential issues. No, we had to recreate real environments with multipath motion interference, traffic loads, and other challenges while monitoring how a given device responds to all these impairments.
To put a finer point on it: Wireless devices and their related software are complex and must be adaptive to various phenomena in open air, such as interference and sudden traffic. Testing these technologies required a control system to replicate real-world challenges and record device responses. This was a multi-dimensional challenge involving very serious RF multipath issues — issues having PhD-level complexity due to the physics from multipath as well as motion emulation and software automation to create different scenarios for testing and recording device behavior.
We ultimately succeeded, and our system became the de facto standard for the whole industry — used by wireless operators, including AT&T, Verizon, France Telecom, and others worldwide. In the process of this massive adoption and serving customers around the globe, I flew enough to get AAdvantage 3-Million Miler status with American Airlines. It was a powerful system, and I wasn’t alone; my cofounders brought in some very smart people. I had to learn how to sell and market it, and that was a new challenge. Then, every year, there seemed to be a different challenge; we became so big that we had to learn how to best grow the team and manage production to maintain the highest quality … things like that. Many very different things must be confronted as a company grows, and we needed to learn something new every year.
What advice would you give your younger self?
Gosh, it always felt like a very anxious and stressful time; I never knew how things were going to end up. So, I would ask myself to chill out a little bit; know that everything’s going to work out one way or another; and understand that getting stressed out never helps. I’d also say just take it easy; think with a clear head; and work as a team member. In fact, that last item might be most important: Nobody can carry the whole load … I’d tell myself that you don’t have to carry the whole load or boil the ocean on your own. There are others, and there needs to be a team.
If you had unlimited time, what global challenges would you be interested in tackling?
I think right now we’re in the age of AI, biotech, and synthetic biology. Networking is done; we did it. Networking technologies are going to keep evolving and getting faster, but their main era of innovation is done. My son just finished college a year ago, and he wasn’t interested in computing. Unlike when I was young (and computers were just coming online and cool), computers now are old news. I tried to get him to do some programming, and he wanted to play with genes and create engineered organisms. So, he went into synthetic biology. This is the age, right?
But there’s a lot of technology in biotech — including computing, automation, AI, and machine learning. We’re kind of still doing engineering and programming, but now we’re working with genomes. People who used to be in engineering, a lot of them are going into biology, and now they’re flipping genes and creating different organisms or trying to fix imperfections in the design of humans because God knows we’ve got a lot of issues we could fix so we could be healthy. On the longer ride, that’s kind of the next wave.
To learn about Miller’s current projects, visit mlinarsky.com.
Read More: The woman who creates engineering companies