Startup

OCTOBER, 2020

Safe Haven

Here and now, I feel safe. This might be due to the effect that this place has on me more than the reality of my situation. Hiding in this old house, with its thick stone walls, double windows and wooden shutters, I feel secure as a mouse in its den.

Outside, I can hear the rhythmic sound of the waves lapping against the stony shoreline, some twenty feet from my window, and I can discern the sea, the mountains and the night sky, all blending into blackness along the Bay of Kotor.

And inside my head, I can also hear the now distant echoes of excited children—the sounds of my cousins and myself romping through my grandparents’ seaside home. But that was more than half a century ago, so far away both in time and experience.

By way of introduction, my name is Andrija, but my Anglo friends, as well as my colleagues and neighbors, find the Serbian name hard to pronounce (it’s that ‘j’ toward the end that always confuses) so they have long ago changed it to Andrew. By now I am quite used to it, or perhaps I am now truly more of an Andrew than an Andrija.

I am originally from what used to be Yugoslavia, but my compatriots managed to screw up that nation, and I was lucky enough to get out before the worst of the bloody disintegration. I was truly fortunate and after getting my EE degree at the University of Belgrade, I stumbled into a scholarship that funded my graduate studies in America. And I have lived there ever since, other than visits to the Montenegro coastline every couple of years to catch up with family.

And if neither Andrew nor Andrija works for you, then call me Professor. Everybody does, because I formerly taught at one of the top engineering schools in the world. But that seems like another lifetime, even though I left academia only a few years ago.

CIRCA 2010:

THE ACADEMIC

Good Old Days

Yes, I was a professor, tenured at a university that many of us believed to be equal or better than the famous Ivy League schools. Holding the coveted Alfred S. Harris endowed chair since 2001—an amazing achievement, if I say so myself, for an academic who at the time had not yet turned forty and with what then seemed like sufficient funding to pursue leading edge research in my chosen field: semiconductor technology. I had a state-of-the-art research lab, which at its zenith was staffed by half a dozen permanent technicians, a couple of associate professors, two or three visiting researchers, up to eight Ph.D students, a handful of MSc grunts, and a number of operators and administrators.

Yes, I was flying high back then. I could happily hand off most of the boring teaching chores to my TAs and concentrate on research. And my team was churning out some excellent work. We were publishing dozens of papers each year that routinely won awards and recognition at many industry events and conferences. Our grant applications were frequently funded by DARPA, or SRC, or NSF or even by various private foundations or corporate programs. Attracting talent was not an issue. The best and brightest were vying to join my team.

Back then I was like a rock star in the business, and invitations for keynote talks, review papers, contributions or simply introductions to various technical books were often extended. I could happily turn down all sorts of speaking engagements, even the private corporate invitations that offered those obscene, but so very tempting, honorarium fees—$10,000 plus expenses for a lecture and a two-hour-long round-table chat—evidently a small price to pay for an opportunity to nourish the egos of a few corporate bigwigs who enjoyed grandstanding in front of a famous professor.

Yes, it looked like the millennium had brought good things for us, and the sky was the limit.

And not just professionally...

Bev and I had met back in the ’90s in one of those combined interdepartmental undergrad classes—something like ‘Science, Technology and Society.’ I got involved with the course because it was trendy, and an easy way of earning an extra teaching credit. A feather in a cap for a newbie, especially because such courses were shunned by the more senior professors who did not have to worry about burnishing up their teaching rep. I thought that teaching science to non-scientists would be easy and would not require much prep work—something that I could easily do off-the-cuff. But as fate would have it, the class held something much more significant than teaching credits. The moment I walked into the first lecture, I noticed her, and everything changed. It was not just the deep azure eyes that were such a contrast to her jet-black hair, or the tight jeans that showed off all her beautiful curves, or... It was the dimples in her cheeks that seemed to amplify the sparkle in her eyes whenever she smiled. And the tiny furrows between her knitted eyebrows whenever she raised a question. And the insidious acuteness of the questions she would raise.

I must admit, with all the brilliance of hindsight, it was lust at first sight—certainly so for me. Ethics be damned!

"But, Professor Krstic, why...?"

Sounding awkward and tongue-tied, I tried to focus on the question rather than on her.

Please explain… With dimples framing a most enchanting smile, her eyes dared me to impress her.

So I had to be stellar in that class, just to keep up with her questions. I mean, how does one explain magnetism or electricity to a non-engineer without sounding stupid or condescending?

But as we got to know each other, the relationship deepened, we fell in love, moved in together, and…well…lived happily ever after. A couple of years later we married. We bought and renovated a perfect house in a good neighborhood. Summer breaks in Europe—often in Montenegro where we congregated with my family. Winter or spring breaks in Mexico. Fall weekends camping in New England… A few years later Lara came, our wonderful daughter, which of course changed everything. All for the good, though. Diapers, pre-school, play dates, kindergarten, school. We settled into a family routine of two professional careers and a kid and looked forward to the continued bliss of middle-class existence in modern America.

__________

But then Moore’s Law caught up with me. Of course, Moore’s Law is not a law of physics but is rather a historical trend that states that the number of transistors in an integrated circuit doubles every couple of years. What became obvious with the brilliance of hindsight was that in order to enable that doubling, the equipment used to make the chips was necessarily becoming more complex. Exponentially so! And the price of that equipment was exploding, too. Certainly, the technological miracles that this equipment enabled suggested that some of the basic laws of physics did not apply anymore. Like printing features many times smaller than the wavelength of light used to define them!

Yes, the late aughts were a time when not only the research labs, but also small, medium, and even full-sized factories—fabs in the lingo—were priced out of relevance, to be replaced by mega fabs: the only ones who could afford the state-of-the-art equipment, which ran into tens of millions of dollars. These Mega-Fabs cost a few billion dollars; an investment that could be amortized only by run volumes of the order of tens of thousands of wafer-starts per month. Literally, acres of silicon! Volumes that could be supported only by standard and commodity products—like general purpose processors or memories. Or that had to be smeared across many IC products from multiple customers—like with the offshore foundries. Tough times even for the industry to fund a team with a dedicated research lab—let alone R&D institutions or academia.

I was discovering that in order to maintain the relevance of my lab, I was spending an increasing amount of time on the road seeking funding for ever newer and more expensive equipment. I stopped being a technologist and instead became a beggar. Not to mention that I stopped being a father and a husband and instead became a road warrior. And to make things worse—adding insult to injury, so to speak—the people whose butts I needed to kiss to get the necessary funding seemed to be getting progressively younger. Maybe it was I who was growing older, but the men that I needed to impress seemed to me to be looking more and more like my students. I found it a bit humiliating to be begging for money from self-important brats who were not smart enough—or humble enough—to know just how little they actually knew. Not something I expected to be doing at my age, having achieved the status for which I’d worked so hard…

Of course I tried to adjust to the changing realities and shifted the focus of my group to the more specialized analog, RF, and even MEMS devices. And then stooped to scraping the proverbial barrel with some of the technologies used for advanced packaging. We shifted toward niches in technology that were the domain of smaller, less sexy, and far poorer sectors of the industry while trying to remain a big fish, albeit in a smaller pond.

But it was all to no avail.

Students—the job whores that they are—adjusted their focus toward arenas that were more likely to get (and keep) them employed. Other departments in my prestigious university—the Computer Science, System Architecture and Chip Design Technology ones—were attracting the cream of the crop. And consequently, the work of my team was becoming ordinary—certainly less stellar. Over time, the only good talent that I could attract increasingly came from abroad: the obsequious Indians; the hairy Persians; the mumbling Chinese; the pushy Koreans. All seemingly blessed with rich daddies and driving shiny BMWs. While I, their brilliant prof, was hiding behind the façade of an absent minded and indifferent academic who could not be bothered to upgrade his clapped-out old Toyota.

The degradation—this slow erosion of our significance—was, of course, quite gradual: so much so that at the time it was almost imperceptible. I certainly was not aware of it on a day-to-day basis. Maybe just on those rare occasions when I ignored the in-my-face issue and took the time to ponder the bigger picture. Like on those long summer weekends when Bev and Lara might have been away somewhere, leaving the house empty. On such occasions I would sometimes allow myself a couple of glasses of bourbon, rather than working late on some paper, or on reviewing the student intake for the coming year, or writing a grant application, or tending to whatever bureaucratic chore-de-jour may have been the pressing issue. Or during some long transoceanic flight when my body was sufficiently confused by jetlag that I could not go to sleep, and just could not face opening my laptop to catch up on all the email traffic with the students, sponsors, colleagues, conference organizers, or whoever else needed an urgent response. Not that these occasions of deep reflection—some might call them navel-gazing sessions—made much difference. I could—and did—shift my attention onto topics that I thought were important. But it made little difference. The industry trends were dominated by factors far larger than me and my lab. I feared that I was just flotsam riding upon an oceanic wave. Of course, most of the time I was successful in pushing those realizations to the back of my mind and carried on being—or perhaps pretending to be—the brilliant professor who had so much to contribute to the field of semiconductor technology.

I first met Aram around that time—around 2012, a few years ahead of the nadir point in the arc of my career as a hot shot academic researcher. He claimed that he had been in one of my undergraduate classes the previous year, but I had not noticed him then. I became aware of him as an individual—as opposed to a faceless student—when he signed up for one of my graduate level courses, because of his name, Aram Khachaturian. Like the composer—that veritable genius who had written several pieces of modern classical music that happened to be among my favorites. When I asked him about it—if he was related to the composer, he explained, Khachaturian is a very common Armenian surname. It was derived from a trade—like Smith in English. He further elaborated that Khachaturian was based on the Armenian word for ‘cross’ and therefore implied that some ancestor of his had probably been a priest or maybe a crusader.

Interesting and notable, I thought, since my own last name—Krstić—has an identical meaning in Serbian. We had a good laugh about that: we were practically twins.

He also stood out in the class, The Physics of Solid-State Semiconductors—one of my most popular ones. Not so much due to the brilliance of his work, which was pretty good, but because he was actually paying attention to the lectures I was giving. Taking notes, asking questions, and actually getting it. Not like most of his fellow classmates with their noses buried in their laptops or phones and only peripherally aware of my existence. I don’t know if the growing malaise of my career at that time was warping my perception, or if the quality of the students—even those at the graduate level—had actually deteriorated, but I did find their indifference quite disrespectful.

Sometime later Aram came to my office and asked about working in my lab. He was looking into doing graduate work on Artificial Intelligence—a topic that has been around for a while but that has all of the sudden become sexy, another of those new solutions— like GPS or Uber or whatever—which were not new in principle but have been resuscitated by the technological advances enabled by Moore’s Law. Artificial Intelligence methods were defined back in the ’50s and ’60s, but the necessary training data sets and cheap compute power became practicable only in the 2010s; hence, the resurgent interest. Regardless, it was a good topic for research. And I told him so.

He said that he did not want his doctoral work to be just ‘pencil pushing’ (his choice of words) and that he hoped to do something hands-on. Something more than just writing code and doing computer modeling and simulations, which was the bread and butter of the design groups. He wanted to focus on an aspect of AI design and architecture that could be demonstrated using the capabilities of my labs. He wanted to build and/or test suitable representative vehicles. He wanted to do something ‘real’ (again his choice of words) not abstract. This was of course music to my ears, and I readily agreed to support, and even to co-sponsor, his graduate work.

Since then I have talked with Aram more or less every day, but back then the interactions with him became the highlight of my days. He was bright and stimulating, eager but not foolish, intelligent, and hardworking—a joy to have as a student and a colleague.

During his graduate work years, we grew close—more so than the usual professor-student relationship. Nothing inappropriate, of course. No, Aram and I just got to the point where we could drop some of the formality in our relationship, and on those nights when we were both working late, we could even have a beer and a bite together. More like friends and equals. I liked him.

It was during one such late night ‘beer-and-burger’ outing that he first floated his idea of starting a company based on our research work. He’d talked before about not wanting to work for someone else, but that was vague; just talk. Not an unusual feeling amongst graduate students who viewed the prospect of a career in the corporate world with a sense of foreboding, no doubt amplified by fear of the unknown, their liberal idealism, and the trendy disdain for ‘the man’.

But for Aram it seemed to be more than just an aversion to becoming another cog in a corporate machine. In his case, it had more to do with family lore, which dictated that working for someone other than oneself was for losers. Apparently, most of his ancestors—going back to their pre-genocide Armenian roots in the Ottoman Empire—had been self-employed businessmen. I come from a family of independent merchants, so a startup company seems like a natural fit for me. Genetic predisposition, I guess. He shrugged.

I suspected that he felt it was the only way that he would gain the respect of his traditionalist Armenian family. I got an impression that they were biased toward the pursuit of wealth rather than acquisition of knowledge. He had a couple of older brothers and a fairly strong ‘big-daddy’ kind of father—each of whom were involved in the family business, which provided building supplies and associated hardware throughout the state. According to Aram, the three of them thought that he was wasting his time by going to graduate school, and they were not shy about telling him so. In fact, it had apparently become a sort of family joke: lazy, bookish Aram who shied away from ‘real’ work.

And I suppose his scrawny, nerdy looks did not help. He was quite skinny—almost skeletal—and full of nervous energy, constantly scurrying about and twitching his stick-like arms.

To me, personally—at least to my face—Aram’s family was always polite; cordial and respectful. Aram occasionally invited me to various Armenian community celebrations: Orthodox Christmas and Easter, Armenian Genocide Day, Independence Day, and the like. So I had a chance to meet his family on a number of occasions. I suspect that being Serbian, and Christian Orthodox, helped, as their rites and traditions, foods and the ways in which they behaved, were not altogether alien to me. So I was not a complete outsider. But neither did they treat me as an insider. They were obviously a tight family who shared their true impressions only amongst themselves.

Nevertheless, no matter how much Aram denied it, I believe that he felt that a startup was the only way he would impress his family and justify his years at the university. "Why would we not do a startup, Professor? he demanded. I’m serious: we’d have to be stupid or blind or both to ignore such an opportunity. We could turn our combined knowledge and know-how into some real money. AI is the up-and-coming thing. I know it. You know it. And we have an AI technology solution like none other." Enthusiastic and eager, he always got intense about making money.

Well, I tempered, because there is much more to it than just saying that you want to do it. You need a good product idea. You need to do serious work to define a business plan. You need to get the seed money by selling the idea to very smart, and very skeptical, people who hate parting with their dough. And then you need to actually develop the best product on the market. And you need to have enough luck along the way to not fail like the nine out of ten startups that end up going under, which, by the way, are staffed by people equally smart and equally motivated as you!

CIRCA 2016:

THE SEED ROUND

Germ of the Idea

Well, Professor? What did you think? Aram asked, plopping down in the chair on the other side of my desk.

But that was just the pretext. Over