“A lot of the time, the risk is still worth taking.”
We’re thrilled to welcome Paul Church, a Google veteran who most recently played a leading role on the Google Cloud Healthcare & Life Sciences team, as our newest Executive in Residence.
Paul fell in love with computers in the mid-1980s while playing educational games at his local library’s summer computer camp. By the time he went off to college a dozen years later, his coding and math skills had turned an early passion into a calling: computer science was an obvious choice.
Through the University of Waterloo’s co-op program, Paul obtained experience in software writing that helped him land a full-time job at IBM in Ottawa. After taking a leave to attend grad school, he was preparing to return when a friend recommended him for a job at Google’s then-tiny Waterloo office. That turned into a 17-year career.
For the first nine years, Paul worked on the ads team, using machine learning to help optimize dynamic bidding, before joining the newly formed Cloud Healthcare team. Most recently, he oversaw the infrastructure that healthcare professionals use to access and store data in the cloud, a job that required close cooperation with international standard organizations.
Paul’s diverse experience at Google means he can support founders on a wide range of topics. Speaking from his 122-acre farm near Waterloo, where he runs a horse-boarding stable, he sat down with us to discuss how he learned to delegate, why people should push past his critical feedback, and how horses have helped him stay sane.
TELL US SOMETHING MOST PEOPLE MAY NOT KNOW ABOUT YOU
In grad school, my research was on tiling theory, which is the study of the way shapes can fit together to cover space using symmetry and geometric transformations. It’s a very obscure field of mathematics, I used to joke that I might be one of the 50 foremost tiling theorists in the world! A year or two ago, there was a breakthrough in this field that made the news, with the discovery of the aperiodic monotile. My thesis advisor was one of the people behind this major result.
CAN YOU SHARE ONE OF THE BIGGEST CHALLENGES OR FAILURES THAT YOU LEARNED FROM?
When working on the healthcare API, there were a number of times when there were features or parts of the product that customers would ask for, and say, “We really need this. It’s very important!” But then, after we spent time and energy building that feature, they didn’t use it.
I realized that customers don’t always really know what they want. What they say they want may differ from their actual needs. So it’s really important to take that surface-level feedback with a grain of salt and to try to dig deeper and understand why they want this.
WHAT’S THE BEST PIECE OF ADVICE YOU RECEIVED IN YOUR CAREER?
“Learn to delegate.”
Twice at Google, I went from being a member of a very small engineering team to being a leader of a medium-sized, and then of a very large engineering team. Both times, the most challenging thing was to be able to let go and delegate to junior members. On small teams, you’re responsible for everything because there’s no one else. As the team grows, you have to realize you have capable people who can take parts of that. Trying to do everything yourself the way you did back at the beginning is not going to work.
WHAT’S A QUESTION YOU WISH PEOPLE ASKED YOU MORE OFTEN?
“Should we do it anyway?”
I tend to approach things with a very skeptical perspective and elaborate on all the risks. So when I’m given a pitch to look at, I often tear it apart, outlining everything that could go wrong, because healthcare is a very challenging space. That leaves people thinking I’m negative on the idea, which is not necessarily the case. A lot of the time, the risk is still worth taking.
WHAT ENERGIZES YOU? HOW DO YOU TAKE CARE OF YOURSELF?
I just go outside, take a walk around the farm, and pet horses. Having a farm kept me sane and also brought me closer to my extended family during the pandemic. They would come visit when there was no school and nowhere else to go.
IN YOUR ROLE AT INOVIA, WHAT QUESTIONS SHOULD FOUNDERS ASK YOU?
Come to me for anything on healthcare standards: it’s a very complex area, with a lot of legacy standards that can be difficult to work with, and I’m happy to share my experience.
I can also help with how to build a product around machine learning or AI. While at Google Ads, only 10% of my work was about fine-tuning models, figuring out machine learning algorithms. The other 90% was about turning the model into something useful. How do you operationalize it? How do you analyze it? How do you monitor your performance? That’s an issue I have an interesting perspective on.