Micro1, a Scale AI competitor, touted to have crossed $100 million

Micro1’s meteoric rise over the past couple of years has catapulted it into a group of AI companies that are expanding at breakneck speed. The three-year-old startup, which helps AI labs hire and manage human experts for training data, started the year with about $7 million in annual recurring revenue (ARR).

Today, it claims to have surpassed $100 million in ARR, founder and CEO Ali Ansari told TechCrunch. That number is also more than double the revenue Micro1 reported in September when it announced its $35 million Series A at a $500 million valuation.

Micro1 is working with leading AI labs, including Microsoft, as well as Fortune 100 companies that are racing to improve large language models through post-training and reinforcement learning, Ansari, 24, said at the time. Their demand for high-level human data has fueled a rapidly expanding market that Ansari believes will grow from $10 to $15 billion today to nearly $100 billion within two years.

The rise of Micro1, and larger rivals like Mercor and Surge, accelerated after OpenAI and Google DeepMind cut ties with Scale AI following Meta’s $14 billion investment in the vendor and its decision to appoint Scale’s CEO.

While Micro1’s ARR is growing quickly, according to the founder, it has yet to match its rivals: Mercor’s more than $450 million, sources told TechCrunch, and Surge’s reported $1.2 billion in 2024.

Ansari attributes Micro1’s growth to its ability to quickly hire and evaluate domain experts. Like Mercor, Micro1 started out as an AI staffing company called Zara, matching engineering talent with software roles before pivoting to the data training market. This tool now profiles applicants and veterinarians looking for expert roles on the platform.

In addition to providing expert-level data to leading AI labs, Ansari says two new sectors, still barely visible today, are on track to reshape the economics of human data.

The first includes non-AI Fortune 1000 organizations that will begin building AI agents for their internal workflows, support processes, finance, and industry-specific tasks.

Developing these elements requires systematic evaluation: testing pilot models, ranking their outputs, selecting winners, fine-tuning them, and constantly checking performance in production. This cycle relies heavily on human experts evaluating AI behavior at scale, Ansari says.

The second is pre-training for robots, which requires high-quality human-made demonstrations of everyday physical tasks. Micro1 is already building what Ansari calls the world’s largest robot pre-training dataset, collecting demos from hundreds of professionals who record object interactions in their homes. He said robotics companies will need huge amounts of this data before their systems can work reliably in homes and offices.

“We expect a significant portion of product budgets in non-AI organizations to go toward human evaluations and data, moving from 0% to at least 25% of product budgets,” said the CEO, who founded Micro1 while at UC Berkeley. “We’re also helping robotics labs generate robotics data; both of these areas will make up a huge share of this $100 billion-a-year market.”

Even as new markets emerge, Micro1’s current growth still comes primarily from elite AI labs and AI-heavy enterprises. The startup is expanding its work with these labs on reinforcement learning, a feedback loop to test and improve model behavior.