Founder’s Lessons: Zoe Liu of Visionular
Zoe shares her experience with video codecs and technology, her transition from research labs to the corporate world, and what makes her unconventional co-founder pairing work.
About
Welcome to the 22nd installment of Tsingyuan Ventures’ Lessons from Founders series. Every week, we publish an in-depth founder interview, ranging from early-stage entrepreneurs to successful businesses. Our conversations cover their personal journeys, the lessons that shaped them, their visions for the future, and their failures. We also learn more about their companies and about the challenges they try to solve. These insights and lessons are applicable to any entrepreneur — current or future.
Read past interviews here.
Visionular is a next generation video encoding and streaming technology software company. They build both cloud based and on-premise video solutions that leverage innovative technologies of machine learning, video processing and compression, targeted to provide the best possible quality experiences to all video users.
Zoe Liu earned her Ph.D., ME, and BE from Tsinghua University and a Ph.D. from Purdue University where her thesis was titled “Layered Scalable And Low Complexity Video Encoding: New Approaches And Theoretic Analysis.” As a prolific IP developer, Zoe has published more than 45 international conference and journal papers and filed over 50 international patent applications.
Prior to founding Visionular, Zoe was a Staff Software Engineer with the Google Chrome Media team for five years where she was a pivotal contributor to the development and finalization of the open-source video codec standard AV1. She has been Principal Contributor/Technical Lead for several renowned video real-time communications products, including Apple FaceTime, Tango VideoCall, and Google Glass VideoCall.
Why we invested in Visionular: Zoe and her cofounder Zheng Zhu formed a world-class video codec (short for “coder-decoder”) team from their lifetime of R&D experience in this area. We are betting on the general tech trend as video is more than 80% of internet traffic today and growing quickly; more efficient compression is worth billions of dollars to video streamers. At the same time, modern video codecs are getting so complex that the AI and advanced optimization based technical moat is significant. Zheng is a serial entrepreneur, but Zoe also impressed us with her ability to learn, scale and build strong trust with customers, partners and other investors.
Meet Zoe Liu
Zoe introduces herself
I’m Zoe, co-founder and president of Visionular. I founded Visionular with my business partner Zheng Zhu in July of 2018. Prior to that, my background was mainly in engineering. I got my first PhD at Tsinghua University in 2000 and came to the United States right afterwards to receive my second PhD from Purdue University. Both of my PhD theses focused on imaging and video codecs. After working with many different research labs, I decided to take on engineering roles at Apple and subsequently Google. I met my co-founder when I was giving a talk for Google in Beijing and from then on, we decided to start Visionular together.
Her beginnings in academia: top at Tsinghua
Back in 1995, I decided to pursue a PhD in electrical engineering at Tsinghua University. At the time, I decided to stay at Tsinghua to continue my graduate school studies, mainly because I was at the top of my class and rewarded the valuable opportunity to be advised by an esteemed professor who was the only Academician of Chinese Academy of Sciences at the time in our department at Tsinghua.
I did not initially start off intending to study imaging and video codecs in particular but arrived at this area of study because of my professors’ guidance. At Tsinghua, my professor was very open-minded and encouraged his students to try exploring all different directions. I studied broadly, such as looking into digital fingerprinting or digital watermarking technologies. That eventually led me to image coding and compression, which seemed really fascinating to me.
On the similarities between research and entrepreneurship
Professors at Tsinghua and at Purdue passed down guidance on the key pillars of research, which I think have similarities to starting a company. In academia, there are three main components: idea generation, experimental results, and presentation. This theory can be applied to industry as well. You must first have a great idea through innovation, then a workable product that customers can benefit from, and finally, a great presentation through marketing. It is not sufficient to have great technology or great products; you must also have a great business team from day one so customers can understand what value you’re bringing. That’s what we try to bring with Visionular.
It is not sufficient to have great technology or great products; you must also have a great business team from day one so customers can understand what value you’re bringing.
Her journey into big tech at Apple and Google
Apple was the first industry job that I held. Up until that point, all of the previous positions I had were in research labs where I mainly focused on patents, papers, etc, but did not really get engaged in the conversion of the research to a real product. Joining Apple as a video engineer and working on building real products was very gratifying, because I had never written code that went directly to a product ready to ship. I worked on streaming for Apple TV and then different versions of FaceTime which was the first video calling application on mobile phones. My final project with them was on AirPlay.
Joining Apple as a video engineer and working on building real products was very gratifying, because I never wrote code that went directly to a product ready to ship.
Even to this day, I would say that this was the most challenging role that I had throughout my professional career (even more so than starting my own company.) This is partly due to the transition from working in a lab, where I was never heavily involved with a product, to the most valuable tech corporation with tremendous pressures on timelines. Apple never fails to deliver on time so there were definitely high expectations. However, it was a very rewarding experience. I learnt a lot about the product cycle where much less time is spent on the software development and much more is focused on debugging and covering corner cases to ensure a seamless launch.
A lot of times, people say failure is the mother of success. Yet, if you constantly experience failure, you will never be able to see what success looks and feels like. For me, the best part of joining a big tech company like Apple was being able to work on a project from start to finish and see the successful results in customers’ hands.
A lot of times, people say failure is the mother of success. Yet, if you constantly experience failure, you will never be able to see what success looks and feels like.
On video codecs and the power of persistence
The biggest lesson I learnt from my time at Google was on persistence. While I first worked on the Google Glass team, I later transferred to a team formed out of a startup acquired by Google to mainly build new video codec standards and develop open source codec solutions.
Let me provide some context for what video codecs are. Every video has to be compressed because videos take a lot of data. We are able to send and share videos and do video conferencing like on Zoom today because of video compression. This process requires some sort of format, which is equivalent to what we call a codec standard.
My team at Google was previously focusing on their own format and not what was mainstream. Back then, it wasn’t popular but they continued working on it and by 2013, they rolled out the next format standard referred to as VP9 which was very successful and adopted by YouTube. As a previously small startup, they never could have imagined these outcomes when they first started (an acquisition by Google, adoption of their format by Youtube, and even more) but their hard work and persistence that made it possible. I carry these lessons with me to Visionular, my own video encoding and image processing technology software company.
On Visionular, AV1 and video codecs
One of Visionular’s core products is the Aurora1 AV1 Encoder. As discussed before, there are many different standards for video compression. AV1 is one such standard which is an open source & royalty-free video codec. It has already become a standard that has been widely supported by Intel, AMD, nVidia, MediaTek, Broadcom, Samsung, etc.
AV1 specifically supports screen content videos (for example, for screen sharing on a video conference). Visionular’s Aurora1 AV1 encoder was found to be the most efficient and performant AV1 Codec Implementation by Moscow State University’s (MSU) annual codec evaluation, which is the global standard for measuring codecs. Using our technologies, we can reduce the data needed for the video by 30–50% while preserving the same video quality. In some instances, we can also improve the video quality while compressing it through our machine learning tech stacks.
The future for AV1 looks bright. Because of the pandemic, video usage has been experiencing astounding growth. We use videos, not only because we watch YouTube, but also for a great range of professional industries. For example, AV1 could have great implications for online education, or even in finance where we’re seeing a lot of insurance companies capturing videos for clients.
Visionular’s Aurora1 AV1 encoder was found to be the most efficient and performant AV1 Codec Implementation by Moscow State University’s (MSU) annual codec evaluation, which is the global standard for measuring codecs.
On unconventional encounters & choosing her co-founder
The first time that my co-founder and I met was in October of 2017 when I went to Beijing to give a talk representing Google. A few months later, we decided to found a company together.
It was definitely an unusual pairing. We weren’t classmates or colleagues, we didn’t share any mutual friends, and we didn’t even work in the same continent. While I spent my whole professional life in the US and majority of it in Silicon Valley, he has never even been here.
Yet it works very well. Both of us have many years of experience in video compression and codecs. While most of my experience was in engineering, my co-founder is a serial entrepreneur and I have certainly learnt from him.
I think the reason why we work very well together is because we share many fundamental values. We always put our team as the number one priority and maintain transparent, open communication. If any one of us has a concern, we always speak up about it. We have established a very strong foundation of trust, which also carries over to our overall executive team and employee base.
We have established a very strong foundation of trust, which also carries over to our overall executive team and employee base.
On building a strong international team
Right now, we are actively hiring and I’d say it’s definitely not easy. Our core employees are video codec engineers. We look for many things but an important one is potential. In this way, we follow the Google style of hiring that I was accustomed to where there is a high bar for technical talent but lots of flexibility about how they can contribute to projects. We also look for diligence and individuals with long term, professional goals. We really aspire to support the professional aspirations of every employee.
Communication is number one to us. Pre-pandemic, I used to fly to China every month to meet with the team there. Right now due to COVID, that’s not possible but I still maintain communication with a lot of other platforms. Just the other day, I counted that I have over 20 video conferencing apps on my computer. Different clients and regions require different platforms so it’s all about having every tool in the toolbox.
When communicating with my team, I always seek to put myself in others’ shoes. Practicing empathy is fundamental because even though we are not in the office or working on different projects, we’re all the same team at the end of the day.
Practicing empathy is fundamental because even though we are not in the office or working on different projects, we’re all the same team at the end of the day.
On breaking her personal records at three marathons
Zoe is an avid marathoner. She has famously qualified for the prestigious Boston Marathon starting with a one-mile slow run in less than 2 years, a great feat. When asked about the parallels between running a marathon and entrepreneurship, she says:
Lots of people will say that running a startup is like running a marathon but I personally think they are quite different. I know where the finish line is in a marathon. However, when you found a startup, you don’t know where or when that exit is. It’s not a certain journey for sure.
I know where the finish line is in a marathon. However, when you found a startup, you don’t know where or when that exit is.
Of course there are some similarities, for example, that both need the spirit of endurance and problem solving. A few months after I finished my first half-marathon race and started to train hard for my first full marathon, I could reach the 15 mile mark but found myself unable to go further. I knew that I had to find a solution and way to work around this, which ultimately ended up being to join a running club. The club was actually made up of a lot of engineers as well. The mentorship and encouragement I received from this club of fellow runners helped me significantly and being able to see my progress over time made me even more devoted.
My biggest takeaway from that would be to learn from the people who surround you. Since joining the running club, I have successfully ran three marathons. I learned from those marathoners in my club and now today, I learn from the people within Visionular and other founders in my community.
To get updates on this series, please follow our publication on Medium.
Follow us on LinkedIn: https://www.linkedin.com/company/tsingyuan-ventures/
Follow us on Twitter: @TsingyuanVC
Tsingyuan Ventures is a $150M seed-stage technology firm. We back technical founders across software, life sciences, and frontier technologies. Learn more about our origin story and our approach here.
Questions, thoughts, reflections? Let us know in the comments below. We’re always looking for great entrepreneurs and early stage ideas, and we’re always interested in having a discussion about venture, technology, and anything related. To see more about Tsingyuan Ventures, please visit our website: tsingyuan.ventures.
