Data Enablers: SigOpt's Machine Learning Tools Help Businesses Optimize Operations

From some guy’s garage in San Francisco, a group of math nerds has grown into one of the world’s finest-tuned machine learning operations. SigOpt now helps optimize fraud detection and investments like beer brewing and even the manufacturing of synthetic rhino horns.

SigOpt cofounder and CEO Scott Clark said that’s because his 14 employees, like the machines they create, never stop learning.

“The thing that unites everyone at SigOpt is our insatiable desire to learn and teach new things,” Clark said. “Everyone at SigOpt is an expert in something and has something to teach or learn from one another.”

For his own part, Clark said he would not have been able to create machine learnings startup SigOpt without learning from the mentors in his life.

“I can truly say I stand on the shoulders of giants,” he said. “I’ve had the privilege to study and do research at many amazing universities and national labs around the world throughout my career. The educators, mentors and advisors at these institutions helped shape me into who I am today.”

PYMNTS caught up with Clark to find out what sparked the grand idea to create startup SigOpt and how the company came to be a juggernaut of machine learning and fraud detection expertise.

PYMNTS: Can you explain the business?

SC: SigOpt is a machine learning startup that helps researchers “optimize everything.” Our technology helps companies in verticals from financial services and algorithmic trading to oil and gas automatically tune the configuration parameters of their machine learning and AI algorithms to arrive at the best possible outcomes.

PYMNTS: How does your business overlap with the payment processing or e-merchant world?

SC: While SigOpt can accelerate modeling and simulation across a wide variety of industries, one of our key use cases is financial services.

Hedge funds and algorithmic trading companies, like Quantopian, use SigOpt to optimize their advanced trading models to make smarter investments and drive better returns. Banking, credit card and insurance companies also use SigOpt to optimize the accuracy of their fraud detection and risk models.

PYMNTS: Can you show me some data or proof points on how the company has helped retailers?

SC: Retailers and consumer goods companies have used SigOpt to find optimal configurations for the models that power their products, coordinate business strategies and help with customer research. For example, SigOpt has been used to help brew a better beer, create an optimal shaving cream and manufacture synthetic rhino horns.

PYMNTS: Can you give the history on the founding and launch of the company?

SC: I got the idea for SigOpt while earning my PhD in applied math at Cornell, where I noticed that the final stage of many research projects was a domain expert hand-tuning a model they built via trial and error, hoping to achieve better results.

After receiving my PhD, I worked at Yelp and developed a black-box optimization engine called MOE to address this problem for optimizing advertising systems. After testing the solution out at Yelp, we joined YCombinator to launch the company in 2014 and bring this solution to experts in every field.

PYMNTS: Looking back since founding, what has been the proudest moment for the organization?

SC: I am the proudest of the team of humble experts we’ve been able to assemble to help bring this powerful research from relative obscurity in academia into the hands of practitioners around the world. We enable them to create better models and products, faster and easier than ever before.

PYMNTS: Can you give me some personnel growth numbers?

SC: SigOpt was started with just a couple of cofounders out of my house in Glen Park. We now have 14 employees who collectively make up some of the world’s premier experts on Bayesian optimization and machine learning systems engineering.

PYMNTS: What has been the biggest hurdle? How did the company overcome it?

SC: Our biggest hurdle was going from a couple guys working out of a garage with a cool way to solve a math problem to convincing some of the world’s top investors to take a bet on us and our dream.

Since then, we’ve been able to raise over $8 million from investors like YCombinator, Andreessen Horowitz, SV Angel and others while building an amazing team to help service the needs of our customers around the world.

PYMNTS: What’s the company culture like?

SC: The thing that unites everyone at SigOpt is our insatiable desire to learn and teach new things. Everyone at SigOpt is an expert in something and has something to teach or learn from one another.

Whether it is one of our platform engineers sharing lessons she learned from working at Facebook,or one of the previous professors we employ as a research engineer teaching an intern about a new machine learning technique, the most common thing I see at SigOpt is the constant transfer of knowledge.

PYMNTS: What is next? What does the future look like?

SC: Looking ahead, we’re focused on continuing to build a great business, introducing Bayesian optimization to more customers and growing our team.



The How We Shop Report, a PYMNTS collaboration with PayPal, aims to understand how consumers of all ages and incomes are shifting to shopping and paying online in the midst of the COVID-19 pandemic. Our research builds on a series of studies conducted since March, surveying more than 16,000 consumers on how their shopping habits and payments preferences are changing as the crisis continues. This report focuses on our latest survey of 2,163 respondents and examines how their increased appetite for online commerce and digital touchless methods, such as QR codes, contactless cards and digital wallets, is poised to shape the post-pandemic economy.

Click to comment