Databricks Layoffs 2026: AI Era Talent Movement

Databricks did not run a layoff in 2026. The company crossed a $5.4 billion annualized revenue run-rate at 65 percent year-over-year growth, raised $5 billion at a $134 billion valuation in February, and is sitting on more than 840 open roles heading into an expected second-half S-1 filing. The “Databricks layoffs 2026” search trend is real anyway, because the AI-era talent movement is happening around the company, into the company, and through its acquisition pipeline. The story is a hiring story written in the shape of a layoff story.

People type “Databricks layoffs 2026” into a search bar for one of three reasons. They saw a Blind thread about a small targeted cut in March and wondered if more was coming. They are sitting in front of an offer letter from Databricks and want to know if it is safe. Or they are a hiring manager watching the data infrastructure cohort reshuffle and trying to figure out where Databricks fits in the supply curve. Three different questions. One answer. It only works if you stop treating Databricks as a peer to Snowflake or Confluent and start treating it as the destination most of that displaced talent is heading toward.

Last updated: May 21, 2026

Gregg Flecke, KORE1. Almost three decades placing technology professionals into financial services, insurance, HR outsourcing, and healthcare IT buyers. KORE1 earns a placement fee when you hire data and AI talent through our data engineering and data science staffing practice. That is the bias to disclose. The angle below leans toward the side of “you should be reading the Databricks hiring story as a competitive supply signal,” because that is what we are watching on the desk this quarter.

What follows is built from public sources. The Databricks newsroom Q1 FY2026 update. CNBC’s February reporting on the $5 billion Series K and the $134 billion valuation. SaaStr’s accounting of the 65 percent growth math. Levels.fyi compensation benches updated May 20. The Bureau of Labor Statistics 36 percent projected growth in data scientist roles through 2033. Reporting on the Confluent and Snowflake reductions earlier this spring. And the candidate flow our data and AI desk has been seeing since the post-Reinvent 2025 Mosaic launch wave landed in mid-Q1. The piece sits inside our broader tech layoffs 2026 pillar alongside Tom’s Snowflake layoffs 2026 analysis and the VMware destinations piece if you want the full neighborhood.

What Actually Happened at Databricks in 2026

The factual record is shorter than the rumor record. There was no all-hands layoff memo. There was no WARN filing. There was no quarterly cost-reduction commitment on an earnings call, because there are no public earnings calls yet. Pre-IPO companies do not have them.

What did happen in 2026, in chronological order:

• February 9: $5 billion Series K closed at a $134 billion valuation. Lead investors included Thrive Capital, Andreessen Horowitz, DST Global, GIC, Insight Partners, and WCM. Two billion of the round came in as debt financing. Reported by CNBC the same day.
• March (rolling): A small number of targeted role eliminations surfaced on Blind and LinkedIn from individual employees. No company-wide announcement. No filing. Best estimate from cross-checking the affected posters is fewer than fifty positions across product management, post-acquisition redundancy from the Tabular and Neon integrations, and standard performance management. Calling this a “layoff” stretches the definition past where reporters typically draw it.
• April 22: Q1 FY2026 disclosure that Databricks had crossed the $5.4 billion annualized revenue mark, growing more than 65 percent year over year, with AI-product revenue alone past $1.4 billion. CEO Ali Ghodsi told the press the company is now free-cash-flow positive.
• April 28: New Sunnyvale office announcement. Databricks said the South Bay R&D footprint will roughly double over the next two years. The current Bay Area headcount is about 1,295 in San Francisco alone, per company data shared with employer-data aggregators.
• Ongoing 2026: 840-plus open requisitions live on the careers site as of early May. The largest concentrations are in engineering (platform, ML, applied AI), field engineering (Sales, Solutions Architecture), and product. Field roles dominate by raw count. Senior platform engineering pays the highest comp.

None of those events is the headline that drives the search trend. The trend is driven by adjacent layoffs. Confluent cut roughly 800 in March. Snowflake cut its entire technical writing org and continues a rolling rebalance under Sridhar Ramaswamy that has touched close to 700 total positions since February 2024. AWS, Google Cloud, and Microsoft Azure all ran data-platform restructurings inside their broader AI reorgs this winter. People search “Databricks layoffs 2026” because they are pattern-matching against the cohort.

That is the field. One winner. Databricks is the only name on the list with net hiring. Everyone else is shrinking, holding flat, or running quiet attrition. The supply side of the AI-era data infrastructure market is reshuffling toward one company. That alone is enough to redraw the hiring map.

The 840 Open Roles, Roughly Mapped

The Databricks careers page is the closest thing to a public dashboard on where the company is investing. The mix changes weekly. As of early May the rough breakdown looked like this. Numbers are approximate and rounded to the nearest ten because postings open and close on a daily cadence.

• Field Engineering: about 290 open. Includes Solutions Architects, Resident Solutions Architects, Sales Engineers, and Specialist Solutions Architects across all US regions plus EMEA and APAC. Heaviest US concentration in Bay Area, New York, Chicago, Austin, Atlanta, and remote.
• Engineering: about 245 open. Lakehouse platform, ML platform, Mosaic AI, Genie, Lakebase, Photon, Unity Catalog, security, and reliability. Bay Area heavy with a secondary cluster in Mountain View and Sunnyvale once the new office opens later this year. Bellevue, Toronto, Amsterdam, Berlin, and Bangalore round it out.
• Sales: about 130 open. Account Executives at every level from SMB up through Strategic, Industry Vertical specialists in Financial Services, Public Sector, Health and Life Sciences, and Manufacturing.
• Product: about 55 open. Heavy on product managers for the Mosaic AI, Lakebase, and Agent Bricks lines.
• Operations, Marketing, Legal, Finance, People: about 120 combined.

The headline observation. Field engineering and sales together represent roughly half of all open roles. That is not the profile of a product company in cost-cutting mode. It is the profile of a category leader in land-and-expand mode against the installed base of every Snowflake, Confluent, and Redshift customer in the Fortune 1000.

Where Databricks Is Actually Sourcing the Talent

The 840 roles do not get filled with people walking off the street. They get filled, in the order our desk has observed this year, from these pools.

Ex-Snowflake field engineers and sales

The biggest single source. Snowflake’s rolling sales rebalance under Ramaswamy has produced a steady drip of senior account executives, sales engineers, and customer success managers with deep enterprise data-warehouse domain knowledge. Databricks hires aggressively into this pool because the buyer conversation is similar, the technical depth is real, and the migration story Databricks is selling against Snowflake plays better in the room when the person delivering it used to be on the other side.

Comp deltas are positive for most of these landings. A senior Snowflake SE making $260,000 to $290,000 total compensation will typically clear $320,000 to $370,000 at Databricks once the equity component is normalized for the pre-IPO upside and the strike-price math is plugged into a realistic IPO scenario rather than the current valuation. Some land at parity on cash with much stronger equity weighting. None land down on the all-in.

Ex-Confluent platform and streaming engineers

The March 2026 Confluent cuts hit hard on senior platform engineers and Kafka specialists. Databricks has been building out the streaming side of the Lakehouse heavily since the Tabular acquisition, and the recent Lakebase launch put real pressure on hiring engineers who can run production streaming workloads. Three of our placements this spring came from former Confluent senior engineers. Two went into Databricks’s own engineering org. The third went to a Databricks partner. The pattern is consistent across what our peers report.

Acquisition-route talent

This is the underappreciated pipeline. Databricks has spent more than $4 billion on acquisitions since 2023 and most of the value has been the founding teams. MosaicML brought the generative AI infrastructure team and turned into the Mosaic AI product line. Tabular brought Ryan Blue, Dan Weeks, and the original Apache Iceberg creators, which is now driving Unity Catalog roadmap decisions. Neon brought the serverless Postgres team that built into Lakebase. Tecton brought roughly 90 engineers and the founding team that ran the real-time feature platform at Uber and turned it into the inference layer underneath Agent Bricks.

Add up the four. That is over 300 senior engineers added through acquisition routes alone, almost all of them with deep AI infrastructure experience and the kind of repeated startup-shipping muscle that takes years to grow internally. Most are still inside the building three to four years later. Retention on acquisition cohorts is unusually high. Equity vesting is part of that. Product autonomy is more of it.

Ex-FAANG AI and platform engineers

Google’s rolling AI restructuring, Meta’s 8,000-engineer reset earlier this year, the Amazon AWS rebalance, and the broader Microsoft Rule-of-70 buyout wave have all produced senior AI and platform talent looking for a smaller stage. Databricks is one of the few destinations with a real shot at running a product at scale without losing autonomy to a layered org chart. The applied AI scientists and ML platform principals in this pool tend to land at Databricks within four to eight weeks of their separation. Most have already met someone on the team. The data infrastructure community is small.

Top-program new grads

Berkeley, Carnegie Mellon, Waterloo, IIT, ETH Zurich, MIT, and Stanford. Databricks has a real new-grad pipeline and offers among the highest L3 packages in the industry. The Levels.fyi page shows L3 starting at $250,000 total compensation, which is above OpenAI’s general L4 entry and is competitive with Anthropic’s standard offer. The fight for top-program new grads is the most visible from the outside and the smallest in raw headcount terms. Probably 50 to 80 placements per year through this channel.

What Databricks Is Paying in 2026

Compensation is the variable that decides whether a candidate accepts or counters, and the Databricks comp story is one of the more aggressive in the industry right now. The pre-IPO equity carries real expected value at the current $134 billion valuation, and the salary bands are competitive at every level. Levels.fyi data, last updated May 20, captures the picture.

Two things to read off the table. First, the median engineering manager at Databricks ($980K total comp) makes more than the median senior engineer at most public peers. Second, the L7 ceiling at $1.65 million plus exists because the equity allocation at this stage of the company is rich and the strike price has been moving with each successive funding round. Some of that paper value is going to compress when the company actually prints public-market shares. Some of it is going to keep climbing because the comparable multiples in AI infrastructure have not collapsed yet. Anyone making a comp decision against Databricks paper should run a few realistic scenarios on the IPO valuation rather than treat the current sticker number as a guarantee.

Field engineering compensation is structured differently. Senior Solutions Architects with five-plus years of enterprise data experience are typically landing $280,000 to $340,000 in total compensation, with the base salary in the $190K to $215K range and the rest in equity and on-target variable. Strategic Account Executives land higher again, often clearing $400,000 in total compensation in the first ramp year if quota attainment holds at the level the comp plan assumes, and the variable component is real because the deal sizes are real. Multi-million-dollar Lakehouse contracts are common in the enterprise segment, and the on-target attainment math on a deal that size scales differently than what the same rep was hitting on Snowflake or BigQuery line items two years ago.

Compare those numbers to what the BLS publishes for the broader category. The BLS May 2024 OES dataset showed software developers at a median annual wage of $132,270 nationally. Databricks pays the L3 entry band roughly twice the national median. That is the gap that sets the talent gravity.

Geographic Footprint and Where the Hiring Lands

San Francisco remains headquarters and carries roughly 1,295 of the company’s total US headcount. That is about 16 percent of the global employee base, sitting in the Mission Bay building and the satellite office space the company has added through 2025 and into 2026.

The big geographic story for 2026 is the Sunnyvale build-out. Databricks confirmed in April that the new South Bay office opens later this year and the R&D headcount in the building will roughly double inside two years. The driver is mostly the gravity well of the Mountain View AI talent pool, with a secondary push from senior platform engineers who would rather not commute into Mission Bay. Most of the senior ML and applied AI hires happening in 2026 are landing in the South Bay rather than San Francisco proper, which is itself a meaningful shift from the headquarters concentration we saw in 2023.

Remote-eligible roles are real but selective. Field engineering and sales positions list remote across most US metros, with Atlanta, Dallas, Denver, Charlotte, Phoenix, and the Pacific Northwest as the most active markets outside the Bay Area. Engineering roles are mostly hybrid in a Databricks office market, with limited fully remote exceptions for principal-level individual contributors.

International expansion continues. Amsterdam and Berlin are the EMEA growth offices. Bangalore is the largest non-US engineering site. Toronto has been quietly growing for two years. Sydney has a small but expanding field team.

The Acquisition Pipeline Has Become a Hiring Channel

Most readers of the technical press already know about MosaicML and Tabular. Two more recent transactions matter as much for the talent question.

Neon, the serverless Postgres company, closed at roughly $1 billion in 2025 and brought in the core engineering team that has now shipped Lakebase, the operational database product Databricks rolled out in early 2026. The Lakebase architecture is recognizably the Neon architecture with Lakehouse integration grafted on. Most of the original engineers are still on it. That is a small cohort, maybe 60 to 80 people total, but it represents one of the deepest cloud database engineering benches available outside the hyperscalers.

Tecton, the real-time feature platform, closed in 2025 with roughly 90 employees coming over. The Tecton team turned into the Agent Bricks inference layer. The founding engineers, who had previously built the Michelangelo feature store at Uber, are now driving the agentic AI workflow side of the product. That hands Databricks something that none of the pure-play AI infrastructure startups can match: a real-time feature platform with a paying enterprise customer base already integrated.

The acquisition pipeline has changed shape in 2026. Earlier deals were technology buys with talent attached. The Neon and Tecton transactions look more like product-team buys with technology attached. The economics are different for hiring teams trying to compete against Databricks for the same profiles. A senior platform engineer in 2026 has a real chance of seeing Databricks acquire her company in the next 18 to 24 months. That changes the calculus on whether to take a competing offer.

What This Means for Hiring Managers Competing With Databricks

Three patterns from our placements this year, useful if you are running a competing search for senior data, ML, or platform engineering talent.

Comp is harder to match than it looks. The Databricks paper value at L5 and above is genuinely large, and the implied IPO trajectory is short enough that candidates discount the lock-up risk less than they used to. A $400,000 cash offer from a public company is no longer competitive against a $500,000 Databricks package with a credible 12 to 18 month liquidity window. If you are not willing to compete on equity, you are competing for the candidates who already have a reason to say no to Databricks. That subset is smaller than most hiring teams assume.

The candidates who say no to Databricks usually have a specific reason. Mission fit beats compensation in this segment more than people predict. Engineers who want to work in healthcare, public sector, defense, or biotech often pass on Databricks because the platform plays at the wrong layer for their interest. So do engineers who want to ship physical product, work on robotics, or stay close to a specific domain like construction or industrial. Lead with the domain story. Databricks cannot match it because Databricks does not have one.

Recruiter speed matters more than usual. Databricks runs a fast loop. Most candidates we place who also had a Databricks interview in flight got the Databricks offer in under three weeks. If your interview process takes six weeks from initial screen to offer, you will lose every contested candidate to whichever competitor moves faster. KORE1’s average time-to-fill on data engineering placements has been running at 14 days this spring, which is roughly the speed required to stay in contention.

One more note for the contract-to-hire crowd. The strongest displaced Snowflake and Confluent profiles will not accept a contract role this cycle. The market has shifted enough that the senior bench is going direct, often with multiple competing offers, often closing in under 30 days. Contract staffing still works well for the migration project tier, where companies need a Lakehouse architect for six months on a defined deliverable. It does not work as well for the strategic platform hires this year.

Common Questions Hiring Managers Are Asking About Databricks This Spring

Are there actually any Databricks layoffs in 2026?

No mass layoff. A small number of targeted role eliminations surfaced on Blind in March, fewer than 50 positions across product, post-acquisition redundancy, and standard performance management. Databricks has not filed a WARN notice and continues to operate with more than 840 open requisitions. The search trend is driven by adjacent layoffs at Snowflake, Confluent, and the hyperscalers rather than by Databricks itself.

Is it safe to accept a Databricks offer right now?

Safer than most pre-IPO offers in the current market, by a meaningful margin. The company is free-cash-flow positive, sitting on a $5.4B run-rate, growing 65 percent year over year, with AI product revenue alone past $1.4B, and the S-1 is expected to drop in the second half of 2026. The realistic risk is not solvency. The realistic risk is where the IPO prices versus the current $134B paper mark, because the equity is most of the package at L5 and above and most of the value is in the public-market re-rating that has not happened yet.

Where is Databricks hiring most aggressively?

Field engineering (Solutions Architects, Sales Engineers) and core engineering (Lakehouse platform, Mosaic AI, Lakebase) lead by raw count. Geographic concentration is San Francisco for company-wide functions, Mountain View and Sunnyvale for engineering, and remote across major US metros for field roles. Bangalore is the largest non-US engineering hub. Amsterdam and Berlin lead EMEA growth.

How does Databricks compensation compare to Snowflake?

Databricks pays higher at almost every senior level, mostly through equity. An L5 Databricks engineer typically clears $400,000 to $620,000 total compensation, where the equivalent Snowflake senior engineer band sits around $320,000 to $440,000. The pre-IPO equity carries meaningful upside risk and meaningful downside risk. Candidates moving Snowflake to Databricks usually take a positive offer with a longer liquidity horizon attached.

Will Databricks layoffs happen after the IPO?

Probably some role reshuffling, almost certainly no mass layoff, based on the post-IPO history of similar high-growth public companies. The pattern in this cohort (Snowflake, CrowdStrike, GitLab) is targeted GTM rebalancing within 12 to 18 months of the IPO once the public market valuation settles. Databricks is more likely to compress hiring growth than to actively cut, given the current revenue trajectory and the AI product momentum.

What does Databricks look for that other employers miss?

Production scale experience and a real opinion on platform tradeoffs. The interview loop selects hard for engineers who can talk through ten-billion-row data scale, query optimization at the parquet level, and the difference between batch and streaming semantics under a real production load. Most candidates who fail the loop fail on depth rather than breadth. Hiring managers competing for the same profiles should sharpen their own technical screens or risk paying senior comp for less senior depth.

Should I expect a counter-offer if my engineer takes a Databricks interview?

Plan on it. Databricks will close fast and the equity component is hard to match. The most common retention scenario this spring has been an accelerated promotion plus a refreshed equity grant, which sometimes works for the engineer who is not entirely sold on the platform pivot. For mission-driven candidates (healthcare, defense, biotech), domain conviction beats Databricks comp more often than salary adjustments will. The retention play is mission, not money.

How KORE1 Works the Adjacent Talent Pool

Our data and AI desk has been actively sourcing from the Snowflake, Confluent, and adjacent data infrastructure layoff lists since February. KORE1 places into data engineering and data science roles, AI and ML engineering, and broader IT staffing across more than 30 US metros. Our average time-to-fill on data engineering placements has run 14 days this spring. The pool of qualified candidates who are not currently in the Databricks pipeline is real, but it requires moving at Databricks speed to land them.

If you have a Snowflake, Databricks, or broader data platform req sitting on your ATS for more than three weeks, or you are planning to compete for senior data and ML engineering hires in the second half of 2026, our recruiters can show you the available bench inside a week. Talk to a KORE1 recruiter about the data infrastructure landings on the desk this month.