AivexaNewsSearch
AI news for builders and product teamsChecked every hour

Meta is losing its longtime head of infrastructure

Collected Nov 11, 2026

Meta confirmed to TechCrunch that Santosh Janardhan, its longtime head of infrastructure and co-head of engineering, is stepping down. Dave Brown, who joined Meta earlier this year as vice president of infrastructure, will take over the role. Janardhan is not leaving immediately: according to a Meta representative, he will stay on for the transition and move into an advisory position on Meta Compute, with his departure expected next year. Bloomberg first reported the change, citing a recently circulated internal memo; that memo reportedly cited the need to pass the baton, and no reason for the move has been given publicly.

Janardhan originally joined Meta in 2009 and is credited in the report with helping the company grow into the large internet presence it is today. Earlier this year he was given responsibility for leading Meta Compute, the initiative through which the company is scaling its data center investments and strategy. That detail is the most revealing part of the shake-up: the person who was handed the infrastructure expansion effort this year is now moving into an advisory capacity on that same effort.

Brown comes from a very different background. Before joining Meta he spent 19 years at Amazon, where his work centered on compute, cloud and AI, according to his LinkedIn profile. That is the profile of someone steeped in the public cloud model, in contrast to Meta's approach of building and operating most of its own infrastructure.

To read the move, it helps to know what Meta's infrastructure organization actually covers. Meta builds its own data centers, designs its own server, networking and accelerator hardware, and runs the systems that carry Facebook, Instagram, WhatsApp and its AI workloads. In practice, infrastructure leadership at a company of this scale is less about keeping sites online, though that is part of it, and more about deciding what to build, where to build it and how fast. With AI training and inference now driving demand, that job increasingly involves power procurement, chip supply, cooling and multi-year construction schedules.

The timing is notable given how large Meta's infrastructure spending has become. The report notes that the company's investment in this area has exploded in recent years as it works to keep pace with rivals in AI. That dynamic is what makes the appointment of someone with nearly two decades of Amazon compute and cloud experience logical on its face: the skill set of planning and scaling compute at very large scale is exactly what an AI buildout demands. This suggests Meta may value cloud-style operational discipline and capacity planning as much as its traditional custom-hardware engineering.

A likely trade-off is continuity. Janardhan has been at Meta since 2009 and has overseen an organization whose systems were built up over more than a decade of organic growth. Handing the reins to a recent hire, even a highly experienced one, means a transition period in which institutional knowledge moves between leaders. The stated plan addresses this by keeping Janardhan around in an advisory role through the transition, but an advisory position is typically less hands-on than running the organization.

It is also worth distinguishing the two hats in Janardhan's title. As head of infrastructure and co-head of engineering he sat across both the physical footprint and the engineering organization. The report does not break out how those duties will now be split between Brown and whoever else holds the co-head of engineering role, so the practical shape of the new leadership structure is not yet visible. Whether the COO-style compute responsibilities and the broader engineering responsibilities stay bundled is a question developers inside Meta will have a direct stake in.

For the wider industry, executive churn in infrastructure roles is often read as a signal about priorities. The fact that the departing leader's memo reportedly spoke of passing the baton implies an orderly handover rather than a rupture, and the advisory role on Meta Compute suggests the company wants his judgment on the expansion plan it assigned him this year. Both are consistent with a company trying to sustain a very expensive buildout while bringing in outside operating experience.

Why it matters: Meta's infrastructure decisions shape the compute available for its AI products and for the developers and advertisers that depend on its platforms, so a leadership change at the top of that organization is relevant well beyond internal politics. For teams building on Meta's AI services, the near-term effect is likely to be minimal, since the transition is staged and Janardhan remains involved in an advisory capacity. The longer-term signal, in my reading, is a buildout run with more cloud-style operational thinking under Brown, whose Amazon background points that way, though the report gives no indication of specific changes to products, pricing or availability. Product and platform teams at other companies should treat this as one data point in a broader pattern: AI capacity planning has become a C-suite-level concern, and the people who run it are increasingly drawn from hyperscale cloud backgrounds.

Read at TechCrunch · AI

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

Santosh Janardhan, who originally joined the company in 2009, will be stepping down from his role.