How Labelbox Helps Data Science Teams Build High-Quality Machine Learning Applications

By Dan Anderson ● April 19, 2019

Labelbox is a collaborative training data platform for machine learning applications so that companies do not have to create their own expensive tools. And the company has the vision to become the default software for machine learning teams for creating and managing high-quality training data just like how GitHub is the default for software engineers.

Recently, Labelbox announced it raised $10 million in Series A funding led by Gradient Ventures — which is Google’s AI-focused venture fund. Previous investors Kleiner Perkins, First Round Capital, and angel investor Sumon Sadhu also participated in this round. Gradient Ventures founder and managing partner, VP of Engineering at Google, and board member of Square Anna Patterson joined Labelbox’s board in conjunction with the funding round.

“Labelbox substantially reduces model development times and empowers data science teams to build great machine learning applications. With the new funding, Labelbox will continue to double down on bringing data labeling infrastructure to the machine learning teams with powerful automation, collaboration, and enterprise-grade features. We’re excited to work with the team at Gradient Ventures and appreciate their support as we scale our business to meet customer demand,” said Labelbox founder and CEO Manu Sharma. “We’re also proud to have incredible investors who have believed in us since the beginning, such as Bucky, Ilya, and Bill from Kleiner Perkins and First Round Capital.”

To build high-quality machine learning applications, it requires superior labeled training data. And 80% of the time spent on developing machine learning data is related to data management — which slows innovation and results in long design-build-test cycles. And machine learning workflows generally do not have standard tooling for labeling data, storing it, debugging models, and continually improving model accuracy. Labelbox solves this problem by offering the best computer vision data labeling and management solution for industrial machine learning applications.

“Labelbox is well-positioned to fuel the industrialization of machine learning across many sectors, such as manufacturing, transportation, and healthcare. In doing so, they will unlock the potential of AI for companies across the globe,” added Patterson.

Last year, Labelbox experienced rapid growth. Some of the company’s customers now include FLIR Systems, Lytx, Airbus, Genius Sports, and KeepTruckin. These companies have numerous machine learning teams using Labelbox for building and optimizing models from aquaculture to autonomous driving.

“The confluence of accessible GPU compute and deep learning technology has paved the way for companies to build production AI systems. The hard part now is teaching machines to think. Labelbox is the key technology to do just that, enabling AI teams to develop new breakthrough products,” explained Labelbox advisor and Research Scientist at Open AI Peter Welinder.

With this funding round, Labelbox is planning to continue building the most comprehensive solution for machine learning teams to create/manage labeled training data for computer vision applications. To make this happen, Labelbox is planning to double its headcount in 2019 and hire additional talent for filling engineering, sales/marketing, and customer success roles.