Connecting the Dots: Domino Data Lab Drops Into Data Science Wave

Startup, hatched by three former colleagues from the world’s largest hedge fund, is riding a wave of interest for its GPU-accelerated platform.
by Scott Martin

Event opportunity: Join Josh Poduska, Domino Data Lab’s chief data scientist, who will be presenting at GTC DC on Tuesday, Nov. 5

As Wall Street was morphing into a game of quants, Nick Elprin, Christopher Yang and Matthew Granade saw something big shaping up on the horizon: a data science wave swelling across industries.

So, the three left Bridgewater Associates, the world’s largest hedge fund, and shortly thereafter started Domino Data Lab, an open source data science platform now making a splash with AI developers worldwide. 

“My co-founders and I built a lot of the internal platforms and technology that those quants used at Bridgewater to do their quantitative research — what the rest of the world now calls data science,” said Elprin, the company’s CEO.

The San Francisco company, a member of the NVIDIA Inception program that helps startups scale, in August landed on Inc. magazine’s annual list of the fastest-growing private companies.

Bridgewater to Domino

After leaving Bridgewater in 2013, the three found that what companies lacked most was an industrialized platform for data science teams, so they started Domino Data Lab to fill the void.

“The experience and perspective at Bridgewater let us see the white space in the market to see what technology and products could do,” said Elprin.

Domino’s software platform automates infrastructure for data scientists, enabling users to accelerate research, deploy models and track projects.

Under Domino’s Hood

A data science supercharger, Domino’s customizable environment provides users with data science tools to speed workflows.

Its Domino Analytics Distribution offers a scientific computing stack for programming in Python, R, Julia and other popular languages. Domino offers access to commonly used interactive tools and notebooks, including Jupyter, RStudio, Zeppelin and Beaker.

Domino also provides deep learning packages and GPU drivers, including access to frameworks such as TensorFlow, Theano and Keras. The platform enables access to any NVIDIA GPUs in the cloud.

“Working with NVIDIA has helped Domino build products that allow our mutual customers to automate deployment of workloads to GPUs,” Elprin said. “NVIDIA Inception has also helped us grow our Fortune 500 customer base through  podcasts and conference talks.”

Customer Domino Effect

Companies are lining up. Red Hat, Dell, Bayer, AllState, Gap and Bristol-Myers Squibb are all using Domino to accelerate their data science workflows.

“Our investment in Domino has really paid off — probably a return around 10x in terms of efficiency of our data science community,” said Heidi Lanford, vice president of enterprise data and analytics at Red Hat, in a video.

Image credit: Photo by Shalom Jacobovitz, licensed under Creative Commons.