Meta has confirmed the acquisition of Conductor, a 2024-founded startup that lets developers run multiple AI agents simultaneously to write and test code. Financial terms were not disclosed. Axios first reported the deal on November 4.
What Conductor Does
Conductor gives each AI agent an isolated copy of a codebase, helps developers verify agent outputs, and then merges validated changes into the main branch safely. The platform supported third-party agents including Anthropic’s Claude Code and OpenAI’s Codex. Before the acquisition, the service had more than 100,000 users, including engineers at Meta, Google, and Spotify.
The startup raised $22 million in a Series A led by Spark Capital and Matrix in March. It was founded by CEO Charlie Holtz and CTO Jackson de Campos.

What Happens to Existing Users
The transition timeline is tight. New user registrations and downloads stopped on November 4. Conductor Cloud goes read-only on November 13 and shuts down entirely on November 20, after which user data deletion begins and refunds go out to active subscribers. Pro subscribers have had paid features unlocked for free in the interim.
The local macOS app will continue to receive model updates and critical bug fixes through early 2027. The company states that user data has never been sold or used to train models.
Where the Team Lands
Holtz confirmed on X that the team is joining Meta Superintelligence Labs, where they will work alongside Nat Friedman and Alexander Wang. TechCrunch reports that Conductor users will see “something new” early next year. Axios sources say Meta plans to use the team’s rapid product development cadence to build the next generation of AI programming tools.
The Broader Context
Meta launched a developer agent called Muse Code in August and has been building programmer productivity tools for several months. On the same day Meta announced the Conductor deal, OpenAI revealed plans to acquire Pierre Computer, another developer-tool startup. Meta frames its direction around enabling developers to manage large numbers of autonomous agents to accelerate software development.
