About Me

I help teams make a bigger impact by teaching and supporting them to do what’s most essential — whether that’s strategy, software engineering, architecture, product management, user experience, technical writing, or team collaboration.

I love working with teams where I can make the biggest difference. Typically that means a combination of social impact potential, uncommonly good mutual fit, and driven yet thoughtful teammates.

Experience

Principal Engineer, Deepcell — 2020 to present

Early employee (around #20) leading major efforts over the years:

• Built active learning and cluster-based tools that let a small team of labelers label at over 5 labels per second, per person. This let us cost-effectively run weekly training iterations when we were small.
• Led Cloud team on architecture & product to build out all necessary instrument integrations and let us start the technology access program only 4 months after project kickoff (and I was taking 1 day a week off for paternity leave for the whole project!).
• Led team that built core data analysis and sorting workflows in the Axon Data Suite, on top of our Human Foundation Model. These tools let customer biologists glean insights and do followup experiments within hours — uncommonly fast turnaround for a 1st-gen biotech instrument.

Co-Founder & Chief Architect, Green Chef — 2014 to 2018 (Acquired by HelloFresh)

Started in a Colorado cabin one weekend. Shifted focus every 6–12 months as the business grew.

• Built the e-commerce flow that scaled from zero to first 50 orders in under a month, then to over 10,000 orders per week in one year.
• Created tools that let the company launch 15+ new products (dishes) per week by letting chefs coordinate their work efficiently and stay in sync with food manufacturing and supply chain.
• Architected the transition to a first item master + scanner-based inventory; improved inventory accuracy from 80% to 98%.

CTO, Esper Ventures — 2013 to 2014

• Rapidly iterated on iOS apps (photo sharing, neighborhood social, ephemeral messaging), spinning up new app prototypes to market often in only a few weeks.
• Developed effective mobile growth channels in real-world communities.

Director of Data Services, Playdom — 2008 to 2012 (Acquired by Disney)

Started and grew the team that built Playdom’s analytics platform. After acquisition, tailored and extended it for game studios in Disney Interactive. More than half of the company across 7 distributed game studios used our tools every day to make better decisions using billions of events and TBs of data per day.

Some of the key outcomes we made possible, through our sister team of business-facing analysts:

• Real-time monitoring tools on the whole event stream that let us understand launches and respond to issues within seconds.
• Web-based collaborative SQL reporting tools, so product managers could get self-serve insights in minutes without waiting for engineers. This inspired parts of the Playdom diaspora to later build tools like Mode Analytics.
• Scalable A/B testing platform that let product managers easily run and analyze multiple parallel A/B tests per game to improve key metrics.
• Clean, enriched user profile data that let us measure user lifetime value to optimize millions of dollars of ad spend each week.

And all of this was built on top of products like Hadoop and Vertica, before they became mainstream.

Education

• Bachelor of Mathematics, University of Waterloo (1999–2003). Computer Science; minor in Pure Mathematics.
• Master of Mathematics, University of Waterloo (2003–2007). Computer Science. Thesis: Static Analysis for Efficient Affine Arithmetic on GPUs.
• PhD candidate, Stanford University (2005–2008). Left on leave of absence to lead the analytics team at Playdom.

Selected publications

• Chan, B., Wu, L., Talbot, J., Cammarano, M., Hanrahan, P. Vispedia: Interactive Visual Exploration of Wikipedia Data via Search-Based Integration. IEEE InfoVis 2008.
• Cammarano, M., Dong, X. L., Chan, B., Klingner, J., Talbot, J., Hanrahan, P., Halevy, A. Visualization of Heterogeneous Data. IEEE InfoVis 2007.
• McCool, M., Du Toit, S., Popa, T. S., Chan, B., Moule, K. Shader Algebra. SIGGRAPH 2004.

Availability

I’m currently working on side projects and open mainly to fractional engagements (2–4 hours a week) on teams that may need my help. I’m prioritizing social impact and non-profits.

In June 2026, or beyond, more substantial engagements are on the table.

What I’m Looking For

I want to speed up figuring out mutual fit, so you or your agents can understand it up front. The criteria that matter most:

Medium-term social impact

Work that can make a major difference in health or productivity for many people, especially those underserved or in need. I want to work mainly on translational problems — where there are still hard challenges, but you can see the impact on real people in weeks to a year.

Foundational research is essential, but I don’t have the patience to wait years for real-world impact.

Uncommon mutual fit

I’m a generalist who brings an unusual mix of experience in engineering, product management, decision science, and team mentorship — across multiple fast-growing startups in different industries.

Great mutual fit means a problem that demands someone with my skills and approach, where they couldn’t do it if they hired someone else.

Get things done ≠ constant grind

Teams I’ve led hit major goals without working long hours. It’s more important to help a team stay focused on the work that matters. I default to around 8 hours a day of work on weekdays. Once in a while I’m happy to jump in and burn the midnight oil when there’s an urgent business need.

Remote or short/infrequent commute

I value being able to be available for my family, sometimes during the day. So if you work mainly in-office, you need to be in San Francisco. Longer commutes (e.g., 1 hour) to the Peninsula are okay 1–2 days a week.

I’m always happy to work remote with anyone from anywhere.

Compensation

If I work with a for-profit, I expect to be fairly compensated for my role and the company’s stage — with openness to discuss taking more equity for less salary at earlier-stage companies.

Volunteer positions for nonprofits are also on the table.

Typical domains

Always open to discussing interesting problems, but based on the factors above, these domains tend to be a stronger fit:

• Biotech
• Healthcare
• Education
• Civic governance
• Social impact
• Climate change

SaaS can fit if it meets two criteria:

• Clear impact on user productivity as we shift to agent-based workflows.
• If you’re not a frontier AI lab, you need to be solving a durable, real problem — e.g., specialized workflows + UX for a specific market segment, difficult data integration, or regulatory hurdles.

These areas tend to be a weaker fit, but they’re not a hard pass:

• Fintech
• Adtech
• Entertainment
• Defense

For example, if social impact is clearly a top priority, I’m interested. I interviewed at Even.com, a fintech that helps measurably improve hourly-wage workers’ ability to manage their cashflow.