Scale AI Rival Invisible Technologies Lands $100M to Power Advanced Data Labeling
Ten years ago, Invisible Technologies was just another startup tackling the unglamorous task of data labeling. Today, it has raised $100 million at a valuation above $2 billion, positioning itself as one of the few companies capable of competing head-on with Scale AI. The funding was led by Vanara Capital, a new investment firm spun out of TPG in its first major move since the separation.
Invisible first gained attention for its role in training the original ChatGPT and has since carved out a niche in complex labeling tasks. Rather than focusing on simple, high-volume annotation, the company built a marketplace of highly qualified annotators with advanced academic and technical backgrounds. This specialization is designed to meet the growing demand for precision as foundation models become increasingly sophisticated.
The timing of its rise is notable. Earlier this year, Meta acquired nearly half of Scale AI at a $29 billion valuation, thrusting the data labeling market into the spotlight. Rivals such as Surge AI, Turing, Labelbox, and Mercor are also competing for share, but the competition underscores a broader shift: high-quality, structured data pipelines are emerging as one of the most critical levers of power in AI.
Under the leadership of Matthew FitzpatrickDealroom has a profile for this one. Try Dealroom →, formerly of McKinsey, Invisible has doubled its engineering team, grown to 350 employees, and reached $134 million in revenue in 2024. Beyond data labeling, the company is expanding into fine-tuning tools, data measurement, and enterprise software for industries including finance, healthcare, and supply chain management. By building specialized tools and expertise that clients integrate directly into their AI workflows, Invisible acts as a critical infrastructure partner in making AI systems more reliable and commercially viable at scale.
Source:
A.M.
