Micro1, a competitor to Scale AI, raises funds at $500M valuation

Micro 1, a three -year -old startup that helps to find and manage human contractors for data labeling and training, a $ 35 million series has raised a fund that gives the company worth $ 500 million. The round was led by and 1 Advisors, Dick Costolo, the capital of an initiative, and former CEO of Twitter and COO Adam Bain were co-founded.

Startup is one of the many companies who are seeking to fill the gap in the data market created by the latest changes involved in the scale AI. After the Meta investment $ 14 billion In the scale AI and its CEO has been appointed, OpenAI and co AI labs Google They have planned Cut bond With startups, their research is likely to be anxious that may end in the hands of Meta. (Scale AI Rejection That it shares confidential information with Meta as part of its partnership).

However, AI labs still need these data services, and startups like Micro 1 aims to pick up the slack.

Ali Ansari, CEO of Micro 1 – who is only 24 -year -old – TechCrunch tells that his company is working with a number of Fortunes, along with the top AI labs, including Microsoft. Ansari said that Micro 1 is now generating $ 50 million in annual revenue (ARR) from Million Millions in early 2021.

It is still far away from the larger contestants like Marqueur, which is producing More than $ 450 million ARR, and enthusiasm, which brings reports $ 1.2 Billion in 2024The However, the growth and adoption of micro 1 between the AI ​​labs seems to be increasing at a healthy rate.

As part of the new fund, Micro 1 AI legal assistant Donotop is adding Bain to its board of directors as well as its founder and chief executive Joshua Broder.

“Truly models are the only way to learn Net New Human Data,” TechCrunch told a statement.

Reuters before Report details Micro 1’s funding attempt.

All these companies – Micro 1, Surge, Marker and Scale AI – AI lab with access to a large base of human contractors who can label data and produce data for AI training. It has become an important service that organizations like OpenAI, Ethnic, Meta and Google have to create cutting edge AI models.

Scale AI first dominated this place, with the initial insight, that it could pay relatively little money for low -skilled contractors around the world to help label data for AI model training. However, Ansari says that AI labs have been transferred in recent years and to improve their AI models, companies now need high quality data labeling from domain experts-such as senior software engineers, doctors and professional writers. The hard part became the appointment of these types of people.

It managed to make Micro 1 its AI recruiter, which is interviewed and veted candidates who apply for one of the contractors in the company, or Ansari calls them, experts. Micro 1 says that Zara has appointed thousands of experts, including professors at Stanford and Harvard, and the company planned to add several hundred more per week.

AI training data market seems to be changing again. Now, many AI labs are interested in working with startups for “environment” development – virtual workspaces that can be used to train in simulated tasks of AI agents. Ansari says that Micro 1 is creating new offers in place of the environment to meet this demand.

Fortunately for startups like Micro 1, AI labs seem to be working with multiple training data suppliers. The nature of the business is difficult for any organization to handle all the data requirements of the AI ​​Lab. This means that at least at least, there is a lot of tour business.