Petuum, Inc., a machine learning (ML) infrastructure platform, has closed $93 million in Series B funding led by a subsidiary of SoftBank Group Corp., with participation from previous investor Advantech Capital. Petuum has raised a total of $108 million, making it one of the best-funded early-stage AI startups.
Despite all the hype about machine learning's potential to reshape business, there are few examples of ML applications built and deployed at any meaningful scale, except within the walls of a few large technology companies. These technology leaders have been locked in an aggressive arms-race for the best talent, and have also created a range of proprietary frameworks, tools and ecosystems for developing ML applications. This industry dynamic has caused significant fragmentation, complexity and skill shortages for would-be ML developers at companies outside this elite group. As a result, widespread ML development has stalled, and the promise of AI is yet to be realized.
Petuum has created a flexible operating system and virtualization interface that makes it easy for a broad range of users to build any type of machine learning or deep learning application at scale, deploy many applications on a wide range of enterprise hardware – including datacenters and IoT – and manage applications and hardware from a single terminal. Petuum's platform is based on years of fundamental ML and computation research, and transforms what would otherwise be skill-intensive, time-consuming and high-maintenance programming into a friendly and intuitive experience for non-programmers. In addition, Petuum cuts through the fragmentation of current ML frameworks and competing cloud infrastructures, by providing an agnostic and standardized approach to application development and deployment. The result is democratized access to cutting-edge AI technology for companies of all sizes, in all industries.
"Our team is excited to solve the problem of AI's high barrier to entry," said Petuum Founder, CEO, and Chief Scientist Dr. Eric Xing. "We believe this technology should be standardized, accessible, and mass-producible, so that all can benefit from AI, ML, and deep learning. We have surmounted many technical challenges as we work toward a standard and universal AI platform, and are grateful to SoftBank for believing in our company's mission to change the future of enterprise AI by providing first-class capabilities at scale to anyone who needs them."
Xing, an eminent AI professor and Associate Head of Carnegie Mellon University's Machine Learning department, founded Petuum in July 2016 with co-founders Qirong Ho, Ph.D., and Ning Li, Ph.D.. Petuum's founders have assembled a team of over 50 world-class researchers and developers in Pittsburgh, PA, the birthplace of AI. Petuum's team has received global recognition for multiple contributions to the field, including the best paper award at the ACM Symposium on Cloud Computing, the AI 100 by CB Insights, and the AI Startup Top 10 by GWC's G-Summit 2017.
With this new round of funding, Petuum will continue to expand its technical and business teams, and focus on deploying PetuumOS in specific industries that have high AI potential, but low adoption, such as manufacturing and healthcare.
"We are firm believers in the value that AI can bring to a broad range of industries," said SoftBank Managing Partner Deep Nishar. "Petuum's work will finally help to unlock that value."
For more information, visit www.petuum.com.