Joining CircleUp: Building a New Game of Chess to Change Investing

Warren Buffett says in his 1989 annual report that “Having a dog teaches a boy fidelity, perseverance, and to turn around three times before lying down.”

Taking Buffett’s words to heart, I wanted to reflect on past experiences and distill my choices for joining CircleUp into three points:.

1. An innovative business model with a strong moat
2. A good challenge in big data and machine learning
3. A super interesting role as Chief Architect

The Business

CircleUp is an investment platform powered by technology. We invest directly in high-growth, early-stage consumer brands––aggregating disparate data from over 200 sources to inform our investing decisions and drive conviction.

Sitting at the nexus of data and the consumer investing has a few very important but tricky characteristics:

1. Inefficient early-stage CPG private markets: The consumer VC ecosystem lacks efficiency, unlike tech. There’s significant demand for capital, but very few entrepreneurs will get the chance when so much of the current path to funding is driven by relationships and connections. CircleUp is working to bring efficiency and a merit-based approach to this market.
2. Uniform business models: The business models of CPG are also very similar. Whether a company is selling dog food, shampoo or water, the business operates in essentially the same way where Company A makes and distributes a product, retailers buy the product and sell it to the end consumer (in some cases the companies sell to the end consumer).
3. Broadly available data: There’s endless data on consumer product and retail companies. And, much of it is public. This is a dream for a data person like me.

The uniform business models and broadly available data of CPG is a perfect recipe for machine learning. It’s what makes Helio, the core platform of CircleUp, possible.

Helio is a collection of continuously refined data sources and algorithms with user friendly applications. The vision is to give…

• an investor the conviction to make an investment
• an entrepreneur the confidence and insights to inform business strategy
• a retailer the quantitative reason to bet on consumer products

With better information comes better decisions, and better decisions yield better outcomes for investors, entrepreneurs, and industry players alike.

The Challenge

A business model is like the rules of chess and data is like the moves in the game. The results of the games can be predicted with uniform business models and broadly available data. It’s like we can make investment decisions like chess decisions.

Chess is one of the best illustrations of machine learning advancements. The game has well defined rules and a strong archive of existing games to learn from (and a computer can generate new games).

Having built a Chess Bot, however, I’d be naïve to think private investing in CPG will be as easy to solve as chess.

Firstly, there’s no rule in any business that’s well defined enough to be used to play out and generate the result of the business, therefore, we cannot generate new business data like we can generate a chess game by playing some possible moves.

Secondly, the most important ingredient of machine learning is data. The data available to understand the world of consumer products is constantly changing as businesses are born and preferences or trends evolve. This creates both a unique technical challenge for our team and a meaningful barrier to entry.

Furthermore, this data is widely available, but it is often unstructured, dispersed, and continuously changing sources. A huge part of the challenge is to build a robust data pipeline that can process this data effectively and make machine learning easier.

Lastly, compared to traditional software development, machine learning is a relatively new domain with evolving development practices. It’s an exciting challenge to build a machine learning pipeline that can encourage collaboration among data scientists, enable reproducibility and reusability of the models, and improve agility and iteration of experiment and deployment

The Role

I have invested much of my career in the investment banking technology domain (at firms such as Goldman Sachs and JPMorgan) and then moved into big data and machine learning (at Twitter and Cruise).

I am excited by the opportunity to combine those experiences and disrupt private investing. As CircleUp’s Chief Architect, I will have the opportunity to:

• Drive recommendations for technology adoption from a total-enterprise perspective
• Drive adoption of integration patterns and technologies, considering data integration and business integration
• Evaluate and identify appropriate technology platforms for delivering the company’s services to third parties
• Work across product, engineering, and data science teams to enable delivery of a robust, scalable and global platform that can be used to deliver market-leading insights and guide investment decision-making to evolve CircleUp’s platform for the future

As a bonus, I will be doing this alongside our two visionary co-founders and a group of incredible technologists, investors, and operators.

2019 will be super interesting, challenging and fun. I look forward to it.