Scale AI Valued at $3.5B After Series D Fundraising Round

Image credit: scale.com

Scale AI is a data labeling startup headed by 23-year old Alex Wang. Its value is now $3.5 billion. The four-year-old startup raised a $155 million Series D funding round led by Tiger Global Management. Based in San Francisco, its customers include Airbnb, General Motors, DoorDash, Nvidia, Toyota, OpenAI, Pinterest, and numerous other major corporations.

Scale AI Services

The company has a visual data labeling software platform that uses automated tools to label text, image, audio, and video data. Companies can use this labeled data to train machine learning models, build AI algorithms, and speed up AI development. Large annotated data sets are the “foundational layer that fuels all machine learning”, according to CEO Alex Wang.

Scale AI Alexandr Wang
Image credit: @alexandr_wang

Alex Wang co-founded Scale AI in 2016 with the mission of reducing one point of friction that has been slowing down AI development—annotating large data sets for AI platforms. Developers send the data via the company’s Application Programming Interface (API). From there, the Scale AI platform produces scalable, accurate training data. This training data can then be used to build machine learning algorithms.

Triple Last Year’s Valuation

The startup’s latest $3.5 billion valuation is more than triple its valuation last year. In August 2019, Scale AI raised a Series C financing round that put the company’s valuation at $1 billion. Importantly, the company reached the milestone of being a “break even” business in late 2020. The company currently has around 300 employees, which does not include the thousands of data-labeling contractors. It intends to allocate some of the Series D financing toward product and market expansion as well as software improvements.

Scale AI views itself as playing an impactful role in accelerating AI industry progress. As Alex Wang describes it: “The potential applications of AI are going to deeply improve the world: scalable medical diagnosis to everyone in the world, massive reduction in driving accidents, goods which are dramatically cheaper to manufacture and transport, and much more. What’s holding us back from building those dreams is lack of infrastructure. While many organizations are adopting AI, they are forced to build their technology stack from scratch, significantly slowing progress.”

The Company Evolves

The company is evolving beyond data labeling to meet the vast demands of AI development. At the beginning of 2020, the startup launched Nucleus, an AI platform that manages the cycle of AI development across teams. Wang explains that Nucleus is like the “Google Photos for machine learning data sets.” The Nucleus platform provides customers an efficient way to visualize and organize large datasets across teams.

Scale AI Plans to Acquire Helia

Scale AI also announced plans to acquire Helia, a startup that provides computer vision and neural network training infrastructure. This software infrastructure enables companies to run AI models on real-time video streams. Helia was founded by former Tesla employees that worked on Tesla’s Autopilot systems. Helia’s expertise in real-time video and neural network training is meant to complement the Nucleus platform.

In describing the strategic rationale behind the Helia acquisition, Wang notes “the one thing that we were noticing across our whole customer base was that more and more customers, even beyond just the self-drive folks, were wanting to do AI on real-time video. And so it was becoming this expertise that we knew just wasn’t going to go away.”

Although Scale AI’s focus on high-quality data annotation and ability to scale machine learning infrastructure puts it at a distinct advantage, the company faces competition. Appen, an Australian company, acquired a startup called Figure Eight in 2019 for $300 million. This acquisition brought together two major data annotation companies. The unified business will be able to process increased volumes of data that supports machine learning and AI. Additionally, Amazon.com offers a service called Mechanical Turk, which breaks down time-consuming projects into smaller pieces, and then outsources those microtasks to many dispersed workers over the internet. These tasks may include sorting and analyzing data sets for machine learning applications.

Autonomous Vehicle Development

Scale AI has played a particularly important role in autonomous vehicle development. Customers such as Toyota Research Institute, General Motors, Honda Innovations, and Lyft have partnered with Scale AI to label large volumes of data. As an initial step, Scale AI uses a standard set of labels for pedestrians, cars, and cyclists for image and video annotation. It also layers in more complex types of data annotation methods such as 3D semantic segmentation.

Customer satisfaction with Scale AI’s capabilities has been strong. Brandon Moak, Chief Technology Officer of the self-driving trucking startup Embark, states that “Scale AI makes it easy to focus on developing leading self-driving technology rather than the operations of labeling data. The scale platform provides the technology and operations to power our machine learning development and is our trusted training data solution.”

Scale AI’s platform has also been a hit among companies specializing in drone and robot development. Its image annotation, Lidar, and video APIs have powered computer vision for drones and robotics using high-quality training data.

It is estimated that the AI market will be worth $44 billion by 2024, up from $18 billion in 2020. As AI and machine learning progress continues to accelerate, Scale AI’s ability to provide essential data services and operate at scale will become all the more relevant.