Navigating Startups and Cutting-Edge Engineering: A Journey of Rewarding Challenges and Growth
Bowei, a serial first engineer at Hive, Nearside, Eraser and more, shares his insights on the value of networking, his start into machine learning, how he bets on startups vs big tech through teams, and how Peer2Peer's unique approach played a pivotal role in his transition to Imbue(f.k.a. Generally Intelligent).
Photo credit: Kristin Cofer
Let’s start with an intro
👋🏼 Everyone, I’m Bowei. I've worked as an engineer in four different startups, taking on diverse roles within various industries over the last seven years. In my most recent position at Eraser, an engineering whiteboard tool company, I was responsible for developing customer-facing features for the web application as a part of a compact 6-person team. Prior to that, I had the privilege of being the first engineer at Nearside, where I helped build their financial platform from the ground up, and spent my early career working in machine learning research and productionization at Gyrfalcon and Hive.
We are so glad Peer2Peer is able to help you find your next opportunity with Imbue! Tell us how that came about.
I learned about Peer2Peer through an ex-coworker, Max, who worked with me at Hive. I was intrigued primarily because of the peer-to-peer referral aspect. I’ve been pretty involved with hiring at Hive and at Nearside, and know from experience that the best job hunting experiences usually come from ex-coworker connections. This platform’s model seems to closely follow that, so I decided to give it a shot. And here we are!
Nice, we’ll dig into that a bit later. What is your origin story with Artificial Intelligence?
I’ve always had an affinity towards mathematics and participated in competitions growing up, which brought me to pursuing a PhD in pure mathematics at Stanford. However, I eventually realized that research wasn't the right path for me. A lot of my peers were pursuing careers in trading or finance, but I wasn’t interested in that lifestyle. I was curious about the startup community here in the Bay since it’s quite vibrant. A few of my classmates referred me to Angellist, where I found Hive, which was Kiwi at the time. I was interviewing with a couple of other companies, but Hive came through with their offer first, and I accepted it as their first data scientist. Just like that, my career officially kicked off in AI.
Why early stage startups instead of MAANG companies?
Early stage startups offer a distinct advantage: the opportunity to choose your own manager, due to the direct involvement of the CEO/CTO in the hiring process. By asking strategic questions, you can gain insights into the company vision, decision-making processes such as the tech stack selection, and their approach to team dynamics. Personally, it took me a few interview cycles to discern my leadership preferences and extract them from the interview process – it ensures alignment with an organization that truly suits you.
Another valuable aspect of startup life is the unparalleled experience and autonomy it offers. While you might not have access to the "big-company" playbook of established best practices, you have a substantial role in decision-making and the chance to witness firsthand the outcomes of your engineering choices.
Seems like you have a track record of picking winners. What advice can you give on discerning which startups to join, especially considering the company is still in early stage and pre product market fit?
I place a significant value on trustworthiness in leadership. For me, this means ensuring that the founders align their words with their actions, both in terms of product development and how they treat their team. It’s often said that culture starts from the top, and that’s something to bear in mind, particularly if you’re considering a commitment of one, two, or even three years.
Predicting startup trajectories can be challenging, as many end up pivoting. While they may have expertise in a certain domain, the specific problem they are solving can easily change due to market forces. On the other hand, people's dynamics at a startup are more resistant to external conditions. Founders remain consistent throughout pivots, and many employees stay even if the target demographic or marketing approach shifts.
Let’s dig into some of your career highlights
I’m grateful to have had my formative years at Hive. As a budding engineer, live collaboration directly with our CTO, Dmitriy, taught me to deliver high-quality code efficiently. His guidance pushed me to think creatively and solve challenges adeptly.
One highlight for me was the Friday team demos to promote and share the ML team’s weekly updates. We were a very intense team solving very interesting problems and we loved the chance to come together at the end of the week to present our work. There was often a last-minute rush, coupled with presentation snafus, but those sessions fostered a tight bond within the team to this day.
As I continued to grow with the company, I took on many roles including a glue engineer supporting specialized hires by building connecting systems, shaping interviewing processes and participating in the decisions on some key hires.
What did you learn about hiring?
Similar to Peer2Peer’s ethos, we relied heavily on our professional networks. In addition to leaning on our networks, we needed a hiring process that could tease out the most suitable candidates from our top-of-funnel. One tactic which worked well for us was allowing individual interviewers to devise and refine their own onsite questions, rather than having a shared pool of questions. We would come to a round table discussion on the same day to debate whether we should make the hire. This approach acknowledged the diversity in how individuals gauge excellence. It sounds controversial, however, if we individually assess candidates using various criteria and still collectively agree that they're an excellent fit, it holds significant weight. Of course, this method also has its downsides – if an interviewer was out, for example, we would lose significant data. However, the varied backgrounds of our team and the close syncing and refinement of our questions led us to a strong overall process.
While this method proved effective for our size at the time, it was labor-intensive and naturally couldn't scale as we expanded.
What made you realize that you were ready for something new?
While I had a great time at Hive, the demanding pace was causing burnout - the teams were getting larger and more specialized, but I felt as though my areas of responsibility and workload weren’t diminishing.
I was talking to another Hive coworker who had left, and he introduced me to his former coworker, Thomson, who was starting a fintech company at the time. The company's mission deeply impressed me, and Thomson’s considerate approach to both the business and employees left a strong impression. I was compelled to be a part of the mission at Nearside.
At Nearside, we onboarded two exceptional engineers, Angel (a former colleague from Hive) and Dwayne (previously my roommate). We had a lot of fun primarily because Thomson gave us full autonomy to make early engineering and framework decisions. True to the startup nature, we trusted each other to ship code quickly without breaking the code but that wasn’t always the case. I fondly remember one time we finished building a feature, shot off a deploy, and went to grab some takeout nearby on Kearny street. Just after we placed our order, we started getting notifications for server errors. Dwayne ended up sprinting back to the office (we forgot to run a prod migration), and Angel and I brought his food back for him and shared some laughs of relief.
Following Nearside's acquisition by Plastiq, and subsequent RIFs, I started job hunting again and found myself at Eraser.io, primarily because of my interest in entrepreneurship.
Eraser offered me a chance to delve deeper into the Product and Design aspects of EPD collaboration, an area I hadn’t extensively explored before. Hive lacked a consumer-facing angle, and at Nearside, I didn’t engage extensively with that part of the business. Joining Eraser became my opportunity to learn from a team that had product design as a core competency which I thought would be useful if I ever wanted to explore becoming a founder myself.
Sounds like you were quite happy at Eraser. What prompted the search?
Well, the interesting thing is, I wasn’t really looking when Max reached out to me via Peer2Peer, but because of my personal connection to him, I decided to check it out.
You had some hesitations initially. Tell us about that.
Yes, this is a lesson on why you should always interview instead of believing what you read on paper. Initially, I wasn’t sure what Imbue was trying to build from their website. This led to further questions like how they measure KPIs or mobilize the seemingly flat team towards a coherent vision. The job description was also pretty broad, and didn’t say much about an infrastructure role, which I was most interested in. I decided to schedule a call with Nic Seo, the Head of Collective Design, anyways, and got my question answered while we were on our call. Turns out, they did have a need for building out the ops/support side of the team.
Lesson here is to always ask, you never know!
What convinced you to seriously consider joining Imbue?
The pivotal moment was at the two-day onsite interview. Phone interviews are important for first impressions, but often fail to provide an accurate sense of the actual work environment. Imbue designed an onsite process where interviewees would experience what it’s like to work there on a day-to-day basis– i.e. working on real engineering tasks. I was able to witness firsthand the team dynamics – how peers collaborate, how meetings are run, and how the founders steer the company’s direction.
Coincidentally, I happened to be present during their weekly Monday roadmap meeting, which significantly boosted my confidence in their approach to planning.
How would you compare Peer2Peer versus other recruiting processes?
Overall the experience was remarkably smooth. P2P facilitated a welcoming introduction, crucial for establishing the conversation's tone with Imbue, managed compensation expectations, and then passed the technical processes of the interview to Imbue. What I liked about Peer2Peer particularly was having access to my referrer, Max, as a sounding board throughout the process. It humanizes the job hunting process significantly and hence, makes it much more delightful. Both P2P and Max were very open to feedback and I really appreciated having some people to bounce ideas off of when I was making the final decision.
I highly recommend it!