Scale AI Valuation: Reaching $3.5 Billion in the AI Boom

Scale AI, a data labeling company established four years ago, has determined that providing the essential tools for the development and implementation of artificial intelligence represents a substantial market opportunity.

The organization, which has developed a platform for visually labeling data utilizing both software and human resources to categorize images, text, audio, and video for businesses creating machine learning algorithms, has secured an additional $155 million in funding. This funding round, spearheaded by Tiger Global, elevates Scale’s post-money valuation to exceed $3.5 billion.

Significantly, Scale has reached the point of profitability and is positioned for continued growth in personnel and expansion into new markets in a financially sound manner, according to Scale’s CEO and co-founder, Alexandr Wang, in a statement to TechCrunch. The company intends to utilize these funds to increase its employee base from 200 to approximately 350 by the end of the following year. (This figure does not include the substantial number of contractors employed for data labeling purposes.) A key focus will also be on entering new markets and enhancing its product offerings and platform capabilities.

Scale initially gained prominence by providing labeled data to companies developing autonomous vehicles, enabling the training of machine learning models for robotaxis, self-driving trucks, and automated systems used in warehouses and delivery services. Established automotive manufacturers like General Motors and Toyota, chip manufacturer Nvidia, and numerous autonomous vehicle startups, including Nuro and Zoox, have utilized its platform.

Recently, Scale’s clientele has broadened to include entities within the government sector, e-commerce, enterprise automation, and robotics. Companies such as Airbnb, OpenAI, DoorDash, and Pinterest are now among its customers. Wang noted that this expansion has gained momentum throughout 2020.

“What became clear, particularly over the last year, is the extensive range of applications for AI,” Wang explained. “We believe we are still in the early stages of this evolution and are preparing to meet the demands as they emerge.”

This preparation involves expanding beyond its role as solely a data labeling service. Earlier this year, the company introduced Nucleus, an AI development platform that Wang characterizes as “Google Photos for machine learning datasets.” Nucleus offers customers a system for organizing, curating, and managing large datasets, enabling companies to evaluate their models and assess performance, among other functions.

“Nucleus represents the foundation of our future direction,” Wang stated. “We recognize that a significant challenge for many of our customers is the lack of a comprehensive set of tools and infrastructure comparable to those available for traditional software development within the machine learning space.”

The company’s objective is to continue developing Nucleus into a complete, integrated platform that empowers a wider range of organizations to leverage AI, Wang added.

To bolster Nucleus, Scale completed its first acquisition, purchasing a four-person startup named Helia. The Helia team, possessing expertise in real-time video and neural network training, will contribute to the development of Nucleus.

“We observed a growing demand across our entire customer base, even beyond the autonomous vehicle sector, for AI applications involving real-time video. This indicated a persistent need for specialized expertise,” Wang said.