On the Record with Zeeshan Zia, CEO of Retrocausal
Hometown: Karachi, Pakistan
Hobbies: I used to like playing cricket and running, but I have 2 kids now, so all my hobby time goes to them.
3 words to describe yourself: hardworking, procrastinator, ambitious
What kind of kid were you? Do you think anything in your childhood helped shape who you are today?
In Pakistan in the 90s, most people did not have a computer at home, but I was fortunate to have one, and my mother enrolled me in a computer programming course in middle school. She also pushed me to compete in programming olympiads in school, and I think both of these things were extremely influential in helping me become a strong computer programmer. By the time I went to college, I had already been to 9 competitions and won several prizes, and this not only gave me a leg up above my peers, but it gave me the confidence in the industry that I might not otherwise have had. Because of these competitions, I am able to look at scary statistics and not be as intimidated as other people. For example, you hear that typically only about 5% of VC backed startups actually make it. That’s a scary number. But my background has given me the confidence to not be scared off by that number. If my mother hadn’t pushed me at the beginning, I would likely have learned the tech skills for computer programming anyway, but I would have missed the confidence piece—and confidence goes a long way in a child.
How do you start and end each day?
In the morning I wake up and get the kids ready for school. I end the day by talking with our engineering team in Asia from about 10pm to 1am each night.
How do you think your background has helped you as a tech founder?
During my time at Microsoft, we shipped out 1 entirely new product and 2 or 3 key features to existing products, and that was a great experience because I got to sit in some very large, all-hands-on-deck meetings and see all the gears turning. One of the products took just one year from the day we decided to ship it until the day it was shipped to the day it entered the market—we never thought we’d see such a quick turn around. The product ended up failing (it was Microsoft’s VR goggles), but getting to be part of the process was a huge highlight to my time there.
Can you tell us the founding story of Retrocausal and why you founded the company?
While at a previous job, we were building head mounted AR for frontline workers, and we had the chance to talk to some of our early, enterprise partners. We learned two important things: 1) they were investing a lot in head-mounted AR, but they didn’t really want to put the goggles on their workers’ heads; 2) AR solutions were very hard to program. Regarding the first point, this was very surprising to us. Upon further investigation, it turned out that about half of the large, enterprise partners really came for the “AI-focused-on-human-worker” piece, not the AR piece. That product has a leading human-focused AI platform, so we thought to create a no hold, easy to use, AI platform that was focused on humans and understanding what humans are doing.
Regarding the second point, we learned that a lot of the early customers had to hire large developer teams just to program AR experiences, and we didn’t want that. We wanted any high school graduate to be able to use our platform and build new experiences. So, it was these two points that led us to found Retrocausal.
Can you talk more about the market opportunity for Retrocausal—what opportunities or upcoming product features excite you the most?
I'm very excited about the intelligence augmentation space and helping human workers become more efficient. The discussion in the industry so far has been about robots coming to manufacturing jobs, but having been a robotics researcher myself, I think we’re still at least two decades away from robots replacing humans. You can spend millions of dollars building highly customized machines to replicate humans on a specific process, but robots cannot match the dexterity and human ingenuity of real people. Our platform helps humans become more efficient.
Additionally, there’s this component of improving manufacturing processes which are still very manual today—manufacturing is still done today the way it was 120 years ago. Right now, in order to improve an industrial process, you can run a Kaizen event and note an improvement down in an Excel sheet, but no one else in the organization can really use that to improve their processes. What’s coming up in our product is to automatically propose to industrial engineers how to improve a process, so instead of going through years of inefficient process and discovering by trial and error, we are building counterfactual reasoning into our AI that automatically infers what will be the best way to improve.
What has been the most interesting or challenging industry to work in so far? Do you find there to be fundamental differences between industries or each company provides its own opportunity?
We’re currently working in manufacturing, and that has been pretty challenging. You have to convince a lot of stakeholders in the sales cycle before a deployment is signed off. At the same time, you learn about a lot of things, and once you deploy, you are there for life. It’s a very “sticky” industry—once a decision is made it’s there to stay. There’s also a “land and expand” sales dynamic in manufacturing, so, once we land a deal, that means that it’s that much easier to get in front of the next customer. We’ve also seen how excited leadership teams are about our product in manufacturing, more so than in other industries like medical training or construction. They are really interested in revolutionizing and digitizing the space.
Do you have a specific leadership style or company culture mantra?
Be polite and give people freedom and independence. If someone comes to me passionate about a new idea, I let them go ahead with it even if I think it needs some revision, because I don’t want to dampen their passion. I really believe that there’s no replacement for passion—not experience and not qualifications—so it’s the one thing I really try to encourage. Our leadership team lets the employees have ownership over projects—from fresh college grads to experienced individuals. Sometimes it doesn’t work out, and we need to readjust, but generally it’s been a great strategy.
You most recently worked at Microsoft, which we all know is a huge corporation—how has the transition to start up life been for you?
It has been a steep learning curve. I think about how I used to run sales calls at the beginning versus how our experienced leader runs them now. At the beginning, we had basically a college-project level website until we hired a product marketer consultant to redo it for us. I’ve been in the computer vision/AI field for 15 years, and I've gotten very comfortable in that zone. Now, I’m starting anew and have to prove myself again. I find that exhilarating. Everyday is a highlight because everyday I learn something new.
In addition to this, there’s a lot more personal ownership in a startup. At big companies you don’t have any ownership—multiple people tell you what to build and how to build it. At Retrocausal, I can choose which project to pursue and which customers to chase. At the same time, there’s no safety net or real hierarchy—I just cleaned our toilet, for example.
Lastly, one very surprising difference to me has been access to customers. Big corporations like Microsoft are essentially 20 different companies combined into 1, and there’s someone from each of these product teams trying to talk to the big customers, like Walmart, for example. In order to do a customer interview with a company like Walmart, you have to go through their Account Executive who will tend to prioritize the existing products that are already producing money. It’s very difficult to get to the customer to discuss a new product. It’s very counterintuitive—you’d think someone at Walmart or Honda would want to speak with a product leader at Microsoft, but in reality, we spent a year and a half just trying to develop basic data sets and problem specs from large enterprises. At Retrocausal, I just pick up the phone and call. It was completely unimaginable to me that you could have more direct access to a large enterprise partner at a startup than at a large company.
What attracted you to or excited you the most about partnering with Differential Ventures?
We really needed help with sales and marketing, I loved that Nick has a background in that space, and he’s really been able to help us understand enterprise sales better.
Are there any life lessons you live by?
I’m not afraid to take risks. I don’t care about cars or houses—I try to approach things with the mentality of “what is there to lose?” I think that’s a good thing to have for ambitious people, particularly entrepreneurs.
You have worked and studied in several different countries. Which has been your favorite place so far and where would you ideally settle down if company geography isn’t a requirement?
The US is the best, hands down. The opportunities here are incomparable, and I spent ten years in Europe, so I know! I understand why all the big companies get made in the US—it leads Europe by a huge margin in deeptech. I also love the “big life” mentality in the States—bigger houses, bigger food proportions, bigger roads. The only “downside” is needing a better social safety net for healthcare, schooling, etc… If you can afford it in the States, it’s better than what you could find in Europe, but I hope things get better here for those who cannot afford it. Ultimately, though, I would love to stay in the US in the long run.
—
Zeeshan Zia is the CEO at Retrocausal, a Seattle-based company that builds computer vision software focused on improving manual assembly work in manufacturing. Before founding Retrocausal, Zeeshan worked as a scientist and shipped multiple AI-first products at Microsoft, NEC, and Qualcomm. Zeeshan holds a PhD in computer vision from the Swiss Federal Institute of Technology, Zurich and led a research group as a postdoctoral scientist at Imperial College London in the early 2010s, before joining the corporate world
Related News
