Meta's Infra Chief Exits in 2027. Most 'Production Engineers' Aren't Meta
Santosh Janardhan leaves Meta in 2027. Here's how to find the infra engineers and directors in his org using @Scale rosters instead of LinkedIn title search.
Meta's head of infrastructure is leaving after 17 years, and the directors and staff engineers in his org now have more reasons to move than at any point since 2022. Finding them is the hard part. Search Meta's own infra job title and most of the results work at oil and solar companies.
On November 11, 2026, Meta confirmed that Santosh Janardhan is stepping down. His replacement comes from outside the company, and the change lands at the end of a year of layoffs and forced reassignments in the same org.
Does Janardhan's exit create a hiring window at Meta?
Janardhan's exit opens a hiring window at Meta, but the leadership change doesn't create it on its own. It is the third disruption to hit the same org in one year, and the window should be widest in Q1 to Q2 2027, after his transition period ends.
Dave Brown takes over infrastructure operations. He spent 19 years at AWS, finishing as an SVP, and joined Meta as VP of infrastructure in July. A Meta spokesperson said "Santosh will stay on for transition and he is moving into an advisory role on Meta Compute," and TNW reports he will stay through the first quarter to help with the handover. Reports disagree on whether Janardhan or Brown carries the CTO of Meta Compute title, so don't put either version in your outreach.
Here is the year as a sourcer should read it:
| Date | Event | What it means for sourcing |
|---|---|---|
| Jan 2026 | Meta Compute announced | Sets the scope of the org you're mapping |
| May 20, 2026 | About 8,000 cut (10%), 6,000 open roles cancelled | Fewer internal moves, more load on whoever stayed |
| Around May 2026 | 20-30% of infra and product engineers reassigned to data labeling | Engineers doing work outside what they were hired for |
| July 2026 | Brown joins as VP of infrastructure | Consistent with a succession planned in advance |
| Nov 11, 2026 | Janardhan's exit confirmed | The path upward for directors resets |
| Q1 2027 | Handover period | The point when people start taking calls |
The reassignment figure comes from Pragmatic Engineer, and it probably matters more than the layoff number. An infra engineer who survived the May cut and was then moved to labeling work has a concrete reason to listen to an offer.
Why an outsider from AWS changes the math
Meta brought Brown in as a VP four months before it announced the succession. That fits a planned handover. If so, directors who expected to move up internally may already know they won't.
The next point is my inference, not something the sources report. New leaders tend to restructure along lines they know, and Brown's lines were drawn at AWS. The directors and staff engineers whose teams built Meta-specific systems will feel that first. Those are the people to have on a list before Q1 ends.
What does Janardhan's org actually cover?
Janardhan's org covers the five areas Zuckerberg said he would keep when Meta Compute was announced in January 2026, and those five areas are the sub-orgs you need to map:
- Technical architecture
- Software stack
- Silicon program
- Developer productivity
- Operation of Meta's global datacenter fleet and network
Zuckerberg's stated goal is that "Meta is planning to build tens of gigawatts this decade, and hundreds of gigawatts or more over time." So the work isn't drying up. People leave because of who runs the org and what they've been told to do. That makes "come build at scale" a weak pitch to someone already working at this budget. The pitch that lands is ownership, scope, and a leader who isn't about to rewrite their roadmap.
Why does LinkedIn title search fail for Meta infra engineers?
LinkedIn title search fails for Meta infra engineers for two reasons. Meta's own infra title, Production Engineer, is shared with oil, solar, and robotics companies, and senior infra leaders almost never call themselves "infrastructure engineers." So title search returns mostly noise at the bottom of the org and nothing at the top.
Production Engineer (PE) is Meta's title for the role Google calls Site Reliability Engineer (SRE). Outside tech, the same title belongs to people in energy and manufacturing.
Refolk's index of US professional profiles shows the gap:
| Query (US, current title) | Total profiles | Top employer in 25-profile sample | Share of sample |
|---|---|---|---|
| "Production Engineer" (Meta's title) | 6,658 | Meta (9) | 36% |
| "Site Reliability Engineer" (Google's title) | 6,641 | Google (3) | 12% |
| Director/VP + "Data Center / Infrastructure Engineer" | 0 | None | None |
The totals come from the full index. The employer shares come from 25-profile samples, so treat them as directional, not representative. The direction is still clear.
If you try to recruit Meta production engineers with a bare title search, roughly two of every three profiles you open aren't at Meta, and most of those are in energy or manufacturing. Adding "Meta" as a current-employer filter helps, but it brings back the other problem: a pool of PEs with nothing to tell you whether they work on the network, the silicon program, or the datacenter fleet.
SRE spread across the industry. PE stayed at Meta.
The two pools are almost exactly the same size (6,658 vs 6,641, a 1.00x ratio). But Meta's share of the PE sample is 3x Google's share of the SRE sample. The likely reason is that SRE became a standard industry title and PE didn't.
This matters for ex-Meta people. A PE who leaves often gets retitled SRE or Platform Engineer at the next company, so searching "Production Engineer" with a past-employer filter of Meta undercounts them. Meta infra people use Meta's internal vocabulary while they work there and industry vocabulary after they leave. For alumni, search SRE and Platform Engineer titles with Meta as a past employer.
At the top of the org, the title search returns zero
Refolk's index returned zero US profiles at Director or VP level titled Data Center Engineer, Data Center Infrastructure Engineer, or Infrastructure Engineer. Senior infra leaders name themselves after their org, not their function: "Software Engineering Director, Network Infrastructure" or "VP Engineering, Core Infrastructure." If you need to source data center infrastructure engineers at director level, you have to know the org names first. LinkedIn won't give them to you.
That is why senior infra searches stall before a single call happens. You reconstruct the org chart by hand, guess at team names, and cross-reference profiles one by one, while the candidates who are actually ready to move get picked up by someone who found them first.
How do you map Meta's infra org from @Scale speaker rosters?
You map Meta's infra org from atscaleconference.com, where every @Scale event page lists each Meta speaker with a bio naming their team. Across a year of events, those bios are the closest public thing to Janardhan's org chart.
Names from the 2026 rosters show how specific the bios get:
- Surupa Biswas, Vice President of Engineering responsible for Core Infrastructure (Systems & Reliability @Scale, June 25, 2026)
- Rajiv Krishnamurthy, Software Engineering Director, Network Infrastructure Group (Networking @Scale 2026)
- Arvind Srinivasan, architect, Network Infrastructure Group (Networking @Scale 2026)
- Omar Baldonado, who gave the Meta keynote at Networking @Scale, where Meta announced it is opening MetaRoCE to the industry
- Faisal Siddiqi, who leads engineering in AI & Data Infrastructure (AI & Data @Scale 2026)
The Systems & Reliability roster also names individual production engineers, and the Networking roster includes a sourcing manager from ISCE (Infrastructure Supply Chain & Engineering). ISCE is a non-engineering part of the infra org, and most recruiters miss it. That is a mistake if your client is building its own data centers.
The mapping workflow
- Pull every 2026 @Scale event page. Systems & Reliability, Networking, and AI & Data are the core three for infra.
- Copy each bio's team phrase exactly as written. "Network Infrastructure Group," "Core Infrastructure," and "AI & Data Infrastructure" become your search strings.
- Treat senior speakers as nodes, not targets. A VP like Biswas is rarely recruitable for a typical req. The value of her bio is that it tells you which org the directors and staff engineers below her belong to.
- Expand outward from each node. Search for profiles that use the same org phrase at director, staff, and PE levels. Pair it with "Software Engineering Director" or "VP Engineering" for leaders, and with "Production Engineer" for current PEs, never the title on its own.
Step 4 is where manual sourcing stalls. You know the org is called Network Infrastructure, but LinkedIn's keyword search can't reliably tell a Meta network engineer from someone who mentions "network infrastructure" in a summary, and it can't cover the switch from Meta vocabulary to SRE titles in one pass. Refolk closes that gap: you describe the person in plain English and get back a ranked shortlist.
When and how should you reach out?
Build your list of Meta infra targets now and start real outreach in Q1 2027, when the handover is finishing. Pitch ownership and stability, not scale.
Specifics on timing and message:
- November to December 2026: Map the org and identify directors and staff engineers whose teams built Meta-specific systems. Don't push for calls yet. Most people will wait to see what the new structure looks like.
- Q1 2027: Janardhan's transition period. Reach out with something specific to the person's work, such as a talk they gave or a system they own.
- Q2 2027: If the restructuring follows AWS-style lines, as I expect, this is when directors who didn't get the scope they wanted start answering.
Don't pitch scale to people spending $35 billion a quarter. Pitch them a roadmap nobody is about to rewrite.
Leave out the layoffs and the labeling reassignments. Everyone at Meta knows about them, and bringing them up reads as opportunism. Mention the @Scale talk instead. "I watched your Networking @Scale session on MetaRoCE" shows you did the work, and it tells the candidate you understand what they built, which a recruiter working from title search can't say.
For clients building their own data centers, run a separate track for ISCE. Its people sit inside an infra org backed by $125 billion to $145 billion in expected 2026 capex, and because most recruiters miss the team, almost nobody is calling them.
FAQ
Is Santosh Janardhan leaving Meta immediately?
No. Meta confirmed on November 11, 2026 that he is stepping down, but he stays through the first quarter of 2027 for the transition and moves into an advisory role on Meta Compute. Expect real movement among his directors around or just after that handover, not in the weeks after the announcement.
Who is replacing Janardhan as Meta's head of infrastructure?
Dave Brown, a former AWS SVP who spent 19 years at AWS and joined Meta as VP of infrastructure in July 2026. Reports disagree on whether Brown or Janardhan holds the CTO of Meta Compute title, so check the latest reporting before you refer to either one by that title in outreach.
How do I recruit Meta production engineers without getting oil and solar profiles?
Never search "Production Engineer" on its own. In a sample from Refolk's index, 64% of US profiles with that title were outside Meta. Combine the title with a Meta org name such as Core Infrastructure or Network Infrastructure. For former Meta PEs, search SRE and Platform Engineer titles with Meta as a past employer, because PEs often get retitled when they leave.
The data behind this
Try it on the search you came here for
Stop building boolean strings. Just describe the person.
Type one sentence. I plan the search, read GitHub, public LinkedIn and Crunchbase records, and the open web as it is right now, and hand back a ranked list with the reason next to every name.
01Describe them
One plain sentence. Role, city, stack, stage, whatever matters to you.
02I read the web live
GitHub, public LinkedIn and Crunchbase records, the open web. Not a database that went stale last quarter.
03You read the shortlist
Ranked, with the reasoning under every name. Open a profile, ask a follow-up, narrow it down.
- Staff backend engineers in NYC who shipped Rust in production
- Series A fintechs in SF under 50 people, growing headcount this year
- Maintainers of fast-growing Rust web frameworks on GitHub
- No boolean, no filters, no seat to buy. One box.
- Read at search time, so a profile updated yesterday counts today.
- Every step visible as it runs, every name with its reason.
500 free credits on sign-up. No card, no demo call. See real searches.