Enter DH

Last updated 2019-06-25 21:00

Do not go where the path may lead, go instead where there is no path and leave a trail.

Ralph Waldo Emerson


Hello, nice to e-meet you. My name is Donghun Lee1 and I am from South2 Korea. That said, most of my friends just call me DH; so can you.

I am a multi-disciplinary software engineer passionate about using data science and machine learning to solve interesting problems in a high-impact space. My research interests have been the application of inverse reinforcement learning and imitation learning to healthcare. I enjoy learning previously unfamiliar things and develop deeper understanding in familiar things; and rarely do I care about the formal boundaries of different fields3.

Below, I shared bits and pieces of my life in a reverse chronological order. My hope is that I nurture this page continuously such that it won’t sound like just another resume material.

The problem space that clicks

I am keenly interested in the problem space of environmental risks and resource management. After extensive research and many meditative nights, I came to believe this is the area to which I would like to make a substantial contribution. Realistic projections like “By 2050, there will be more plastic than fish in the world’s oceans.” really appall me. From the early death of my father who was a theoretical physicist I realized my time on Earth is limited. As a corollary, I am determined to allocate my time and talent to building a system that would make our economies more sustainable.

Ecology of Commerce Natural Capitalism

Riding at a graduate school

I studied computational science at Harvard University (M.Eng.). I learned about beautiful mathematical and statistical theories that underpin Machine Learning and data science4. I also worked on various projects from building a neural-network-based self-play Poker agent to building a tool that helps credit analysts intrepret black-box algorithmic decisions and evaluate their fairness5. As an avid gamer who used to play Starcraft competitively, the natural choice for the research area has been reinforcement learning where games are a primary application domain.

Concretely, I have written peer-reviewed papers and a thesis on the problem of imitating and explaining experts in healthcare using Inverse Reinforcement Learning.

The first business that started by accident

My friend and I inadvertently6 started DHJM that has grown to a profitable business ($3M annual revenue). As I write this, it is going strong with a team of over 20 people. The business provides digital sales and marketing solutions tailored to interesting niche products like food waste recyclers that cost $500 or air purifiers that cost $3,000. We came to understand the pain points of high-quality yet expensive hardware manufacturers and offered deals that allows them to focus on manufacturing at a low risk level while giving us a high upside.

My day-to-day involvement with the company was as short as one year, for I had to leave to pursue the masters study in the U.S. As a first-time entrepreneur, I learned so many things that I could not learn from any schools, companies or talks of other entrepreneurs. Most importantly, I learned to appreciate the usefulness of never-give-up mindset and experiment-and-adapt mindset. When we launched our first customer’s product, the expensive food waste recycler, nearly all of the industry experts in a good spirit warned us “Boys, this ain’t the type of business young guys would want his feet on.” They were right for the first 2 months, but we persevered iterating on the tactics until we just figured out how to make things happen. We saw the sales increasing around 20% week to week and became number one in the product category. With the promising results, our fist customer ended up being acquired by a publicly listed company. After the initial success, other companies started coming to us, even before we reach out to them. That was a magical experience.

Finland, my second home


Oh boy, so many fun memories in Finland. At first, I went to Finland as an exchange student and through a series of serendipitous events it just so happened that I got settled in Finland7. To be blunt, I was not the typical student who puts one’s utmost efforts into studies. I always had a good level of curiosity but then, I just thought there are more curious things out of the lecture rooms. The program I was optimizing during college was to minimize efforts into studies, subject to the expected grades higher than X. Instead, I spent most of my free time working on independent projects building apps.

Day for Failure in Helsinki.

I co-organized startups-related events, as the first Asian member of Aalto Entrepreneurship Society whose presence got so big that the venue of the organization became one of the go-to places for presidents and prime ministers visiting Finland.

Startup life..

I co-managed Startup Lifers that connects talented Nordic tech, design and business students and recent graduates with promising startups in San Francisco. I designed a new program called, Startup On Demand, and during my term, our team sent in the largest number of student interns in the history of the program. I studied Business Technology & Computer Science (minor) at Aalto University (BSc) in Finland.

Dinner with Mårten Mickos, CEO of Eucalyptus Systems and former CEO of MySQL.

Gig at Silicon Valley

I worked at Plivo and was responsible for analytics/growth and front-end engineering.

Y Combinator Hackathon

The long tail of miscellaneous facts about me

  • I’m a sociable geek: people who don’t really know me tend to think I am an extrovert. But the trust is that I am just a shy guy who had to learn to socialize.

  • This is how I looked at 10.

Probably 10 or sth.

  • This is my motto calligraphed.

my motto calligraphed: excerpts from Chapter 10, Tao Te Ching

  • I can talk about these for 2 hours without stopping: real-estate investing (residential and commerical), small-cap value investing, sharing economy, Korean Son Buddhism, Eastern philosophy, behavioral economics, Korean & Japanese manga, puzzle games.

  • People I will pay a million dollars without a moment of hesitation for a 1-hour meeting: Ricardo Semler, Ray Anderson, Bruce lee, Masayoshi Son, Il-han Yu, Ku Kim, Yakyong Jeong, Richard Feynman, Charlie Munger, Edwin Land, Will Wright, Lao Tzu. (Not that I have a million dollars now :-))

  • More about me:
  • I love indie games. Especially, indie puzzle games. Some of my all-time favorite games are: Braid, Stephen’s Sausage Roll, The Witness, Kami.

Braid. Stephen's Sausage Roll.

The Witness. Kami.

  • Being from South Korea, I grew up playing a lot of games including Starcraft.

  • I have been a huge fan of Nintendo games. It’s dangerous to go alone. Take this! If you know what I mean.

  • My life axioms follow from Seon buddhism.

Only don't know. Just go straight.

Seon Master Seung-Sahn Sunim

Son Master Seung Sahn.

Things I like

  • The Unreasonable Effectiveness of Mathematics by R. W. Hamming
  • Work Hard by Terence Tao
  • I personally don’t believe in business books whose topics authors have never practiced. It sounds like too generous a filter but it is not. This is one of the reasons I have a weak spot for reading biographies. Here’s one: Maverick by Ricardo Semler.
  • How can I make smarter decisions? “Invert, Always Invert” said the mathematician, Jacobi. The dual of the problem would be to minimze the number of stupid decisions. So we just need to work to become more aware of our own vulnerabilities and work around them. I learned this from Psychology of Human Misjudgment by Charlie Munger.
  1. My Grandfather paid a name-maker a good sum to come up with an optimal name for me. The name-makers, then, would use obscure algorithms to produce names based on babies’ birth timestamps. I heard this was the thing back then. Funnily, Donghun is one of the more common names in Korea.

  2. Not to be confused with North.

  3. Charlie Munger said “If you skillfully follow the multidisciplinary path, you will never wish to come back. It would be like cutting off your hands.” And I agree.

  4. Some of the courses I took: Machine Learning, Convex Optimization, Reinforcement Learning, Statistical Inference, Numerical Methods, Stochastic Methods, Statistical Learning Theory, Real Analysis.

  5. https://www.seas.harvard.edu/news/2018/06/fair-and-square

  6. It started as a small business project. My primary motivation was to earn enough money to finance my expensive education in the U.S. I took things more seriously after we raised a seed round and began to hire people. I ended up sending a letter to Harvard admissions committe asking for deferral; this offered me a year-long runway of taking the business off the ground to profitability.

  7. http://www.dhfromkorea.com/boredom-is-powerful/