Micro1’s rapid rise over the past two years has propelled it into a cohort of AI companies that are moving at breakneck speed. The three-year-old startup, which helps AI labs recruit and manage human experts for data training, 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. This figure is also more than double the revenue Micro1 reported in September when it announced its $35 million Series A at a valuation of $500 million.
Ansari, 24, said then that Micro1 works with leading AI labs, including Microsoft, as well as Fortune 100 companies that are working to improve large language models through post-training and reinforcement learning. Their demand for first-rate human data has fueled a rapidly expanding market that Ansari predicts will grow from $10 billion to $15 billion today to nearly $100 billion within two years.
Micro1’s rise, along with that of larger competitors such as Mercor and Surge, accelerated after OpenAI and Google DeepMind reportedly cut ties with Scale AI following Meta’s $14 billion investment in the vendor and its decision to hire Scale’s CEO.
Even though Micro1’s ARR is growing rapidly, according to the founder, it has yet to match its rivals: Mercor has brought in more than $450 million, sources told TechCrunch, and Surge has brought in $1.2 billion in 2024.
Ansari attributes Micro1’s growth to its ability to quickly recruit and evaluate domain experts. Like Mercor, Micro1 started as an AI recruiter called Zara, combining engineering talent with software roles before moving into the data training market. This tool now interviews and reviews candidates seeking expert roles on the platform.
Beyond providing expert-level data to leading AI labs, Ansari says two new segments, still barely visible today, are poised to reshape the human data economy.
Techcrunch event
San Francisco
|
October 13-15, 2026
The first involves non-AI-native Fortune 1000 companies that will begin creating AI agents for internal workflows, support operations, finance, and industry-specific tasks.
The development of these agents requires systematic evaluation: testing pioneering models, evaluating their results, choosing the winners, refining them and continuously validating performance in production. Ansari argues that this cycle relies heavily on human experts evaluating AI behavior at scale.
The second is robotics pre-training, which requires high-quality, human-generated demonstrations of everyday physical tasks. Micro1 is already building what Ansari calls the world’s largest robotics pre-training dataset, collecting demonstrations from hundreds of GPs recording interactions between objects in their homes. Robotics companies will need large amounts of this data before their systems can operate reliably in homes and offices, he said.
“We expect a good portion of non-AI companies’ product budgets to be dedicated to human assessments and data, growing from 0% to at least 25% of product budgets,” said the CEO, who founded Micro1 while at UC Berkeley. “We also help robotics labs create robotics data; these two areas will represent a massive share of this $100 billion per year market.”
Even as new markets emerge, Micro1’s current growth still comes primarily from elite AI labs and AI-intensive companies. The startup is expanding its work with these labs on reinforcement learning, feedback loop to test and improve model behavior.
Micro1 hopes its early entry into robotic data and enterprise agent development, in addition to scaling its specialized RL environments, will help it capture additional market share as the data war intensifies.
For now, Ansari says the company is working to scale responsibly, pay experts well and keep people at the center of an industry built on training machines.
The company currently manages thousands of experts in hundreds of fields, ranging from highly technical fields to surprisingly offline disciplines. Many earn nearly $100 an hour, according to Ansari.
“There are Harvard professors and Stanford PhDs who spend half their week training AI through Micro1,” Ansari said. “But the biggest change is in the volume and range of roles. It’s expanding into areas you wouldn’t expect for language pattern training, including offline and less technical areas. We’re very optimistic about the direction this is taking.”
