How UCSC alumni are using enzymes to make factories obsolete

Pavle Jeremic understands the difference between abundance and scarcity. His parents emigrated from the former Yugoslavia. He grew up in Boulder, Colorado and later Davis, California. Going back and forth from Serbia to some of the richest towns in the U.S. sparked a lifelong obsession with fixing inequality.

“I’ll put it bluntly,” Jeremic said. “When you can no longer provide for your family, it’s much easier to commit atrocities. So how do you make that line impossible to cross? Scarcity was the problem.”

His company, Aether, evolved out of this sentiment. They have an enormous ambition: ending poverty completely by making it possible for anything to be manufactured at almost no cost. They’re planning to do it by reverse engineering basic biology to manufacture molecules (and they’re doing it with a little help from machine learning).

“Early on, I realized two things: one, over billions of years, nature evolved incredibly sophisticated machinery and it was clear that the solution was there, but, two, the problem was that humans suck at engineering biology—engineering isn’t even the right word, most of the time it’s more like an artisanal craft that doesn’t work half the time instead of building a car or computer circuit,” Jeremic said. 

Humans might not be able to engineer biology, but they could create algorithms that could effectively take the human out of the loop.

“Lots of companies say they use machine learning and biology,” Jeremic said, “But there’s a big problem: there’s not enough data.” 

Machine learning generally applies an algorithm to a large, well-sorted database of information and uses that information to optimize itself (say using a database of reviews to make increasingly accurate movie recommendations). Jeremic says that the lack of data is such a problem that your essentially restricted to using the first three chapters of the machine learning textbook. 

“Most biology-machine learning companies are really just doing linear regression,” Jeremic said. “Which is great and capable of doing important things, but we wanted something more sophisticated. We decided to rethink the entire manufacturing process and build it around the idea of automating throughput.”

As of November 4th, Aether have generated approximately 50 percent of the world’s biological information.

Jeremic toured existing biotechnology companies and realized that many of his potential competitors were little more than well-appointed academic labs. Utilization of automation was low—successful manufacturers try to have machines running as much as possible—and they weren’t taking full advantage of automation. They were focusing on replacing functions that humans were doing rather than taking full advantage of what automation can do, and allowing automatic processes to evolve.

“I founded this company this company with the idea that let’s think about experimentation and analytics, not from the perspective of automating what a human can do but what you could do if you knew the protocols and could automate from there.”

They received their first round of seed funding in 2018 ($500K), having proven that they could generate data faster than anyone else. They received a second round of funding ($12mm) in 2019 to construct a robotic factory that according to Jeremic could be called the highest throughout biological factory on earth.

In the near-term is goal is to help other companies improve efficiency. “Say you’re XYZ big chemical company and you need to make a molecule—we would specially design a factory just for you. Longer term we want to build the factories of the future. We imagine it would look a bit like a shipping container. Say someone in New Delhi needs certain inputs and outputs, we would send them a container that was capable of doing that. The idea is to completely decentralize the means of production and become a society that is completely about molecular design.”

At UC Santa Cruz, Jeremic learned his chops in the bioengineering program, particularly working with the iGEM project. 

“I think iGEM is the most effective and entrepreneurial project on campus, and I think that’s the most effective way to get people to become entrepreneurs, that is to throw them into competitions and give them way too little time to do anything”

His iGEM team was working with Professor David Bernick to recycle the almond waste into a zero calorie sweetener. Jeremic said the experience was particularly good for learning how to manage people and time.

He also credits CITRIS with being enormously helpful. “They were amazing: always available whenever I had questions and they were all experienced. They were the first ones to show us what we actually needed to become a startup: for example we need lawyers, that we needed to incorporate.”

Ending scarcity is an audacious vision for a tiny company but Banana Slug engineers tend to wriggle out of their normal operating constraints.