AI data labelling startup Handshake has acquired Cleanlab, a data label auditing company, in a talent-focused deal. The acquisition adds nine Cleanlab employees, including three MIT PhD co-founders Curtis Northcutt, Jonas Mueller and Anish Athalye, to Handshake's research organisation. Founded in 2021, Cleanlab had raised $30 million from investors including Menlo Ventures and Bain Capital Ventures. The startup developed algorithms that flag incorrect data without human reviewers. Cleanlab received acquisition interest from multiple AI data labelling companies but chose Handshake because competitors like Scale AI and Surge use Handshake's platform to source human experts. Handshake, valued at $3.3 billion in 2022, serves eight major AI labs including OpenAI and is targeting annualized revenue in the "high hundreds of millions" this year.

Handshake acquires Cleanlab in strategic AI data labeling talent acquisition. Handshake originated in 2013 as a service focused on recruiting recent college graduates. Approximately a year ago, the company expanded its operations by establishing a human data labeling division designed to support companies building foundational AI models. Meanwhile, Cleanlab was founded in 2021 as a software startup specializing in tools that enhance the quality of data generated by human labelers. The primary objective of this acquisition is to secure talent, making it essentially an acqui-hire. The transaction will bring nine key employees from Cleanlab into Handshake's research division. This group includes the startup's three co-founders, all of whom hold PhDs in computer science from MIT: Curtis Northcutt (pictured above), Jonas Mueller, and Anish Athalye. Specific financial terms of the deal have not been made public, though it is worth noting that acqui-hires can sometimes result in substantial financial outcomes for founders. Cleanlab had secured a total of $30 million in funding from investors such as Menlo Ventures, TQ Ventures, Bain Capital Ventures, and Databricks Ventures. At its height, the startup employed more than 30 people. The team at Cleanlab are specialists in creating algorithms that can identify incorrect data entries without requiring a secondary human review. This expertise is expected to significantly boost the quality of the data Handshake supplies to AI laboratories. "We maintain an internal research team that continuously evaluates where our models have weaknesses, what kind of data we should be generating, and how to ensure its high quality," a representative noted. "The Cleanlab team has dedicated years to solving these precise challenges." Curtis Northcutt, the CEO of Cleanlab who is recognized for pioneering automated data labeling audits, mentioned that the company attracted acquisition interest from other firms in the AI data labeling sector. However, Cleanlab opted to join Handshake because, as Northcutt explained, competing data labeling companies like Mercor, Surge, and Scale AI regularly use Handshake's platform to recruit specialized professionals - including doctors, lawyers, and scientists - for their labeling projects. Handshake, which achieved a valuation of $3.3 billion in 2022, was projected to reach an annualized revenue run rate of $300 million by the end of 2025. Recent reports indicate the company is now on course to achieve an ARR in the "high hundreds of millions" this year. To date, Handshake has supplied data to eight leading AI labs, one of which is OpenAI.