On June 16, 2026, Ali Ghodsi stood at the Data + AI Summit in San Francisco and disclosed that Databricks had crossed $6.9 billion in annualized revenue, growing faster than 80 percent year-over-year — the fastest sustained growth trajectory in enterprise software at this scale. That number, combined with a $134 billion valuation secured in December 2025's Series L and AI products alone clearing $1.7 billion in annualized run-rate, has done something measurable to Databricks' employer positioning: it has turned an enterprise data company into one of the most sought-after AI platform employers in the United States, competing directly for the same ML engineers that OpenAI, Anthropic, and Google DeepMind are chasing. With 840-plus open roles as of mid-June and an estimated headcount that has grown from roughly 9,400 in 2024 to between 12,000 and 15,000 today, the H1 2026 hiring machine at Databricks' Mission Bay headquarters is running at a velocity its pre-DBRX self could not have supported.
Databricks Headcount and H1 2026 Hiring Numbers
Databricks entered 2025 with approximately 9,400 employees. CEO Ali Ghodsi publicly committed to hiring 3,000 people in 2025 alone — a target that, per workforce intelligence aggregators including Unify and Revelio Labs, the company largely delivered on, with headcount crossing the 12,000 mark by Q1 2026 and tracking toward 15,000 by year-end on the current run rate. That 24-percent-plus annual growth rate is operating on top of a base that had already doubled in the prior two years.
The open-roles figure as of mid-June 2026 stands at 840-plus positions, per ENTRA's tracking of the company's public careers board and LinkedIn activity. The concentration is not uniform. Field engineering — Solutions Architects and Sales Engineers who deploy Databricks' Lakehouse platform inside enterprise client environments — leads by raw count, reflecting the $6.9 billion revenue base that requires proportional customer-facing coverage. But the engineering-side opens are where the competitive intensity is highest. ML platform engineering, Mosaic AI research and inference, Unity Catalog governance infrastructure, and Lakebase — the company's new serverless Postgres database announced at the Data + AI Summit, now handling 12 million database launches per day — are all actively expanding their core teams.
The geographic split is specific. Most senior ML and applied AI hires in H1 2026 are landing in Databricks' South Bay facilities — Mountain View and Sunnyvale — rather than the Mission Bay headquarters. The company confirmed in April 2026 that a new South Bay engineering office will open later this year, with R&D headcount in that location expected to roughly double within two years. For candidates choosing between Databricks' San Francisco presence and its South Bay build-out, the distinction is not cosmetic: the South Bay campus is where the foundation model and inference engineering work lives, while Mission Bay houses the product, go-to-market, and enterprise architecture functions.
What Databricks Pays: Compensation Benchmarks in H1 2026
840 open roles at a pre-IPO company with $134 billion in private valuation creates a compensation structure that does not look like enterprise software any more. Per Levels.fyi aggregates and ENTRA cross-referencing with Glassdoor and Blind data through Q2 2026, the bands by level are:
L3 (entry-level SWE): $175,000–$240,000 total comp. L4 (mid-level): $240,000–$330,000. L5 (senior): $330,000–$480,000. L6 (staff): $480,000–$700,000. L7 (principal): $700,000–$1,000,000-plus.
ML engineers commanding the Mosaic AI team's attention sit at the top of those bands. Senior ML engineers (L5) with foundation model or LLM pre-training experience are clearing $400,000–$600,000 total comp; staff ML engineers (L6) on the Mosaic AI team are in the $600,000–$800,000 range per candidate disclosures on Blind and Levels.fyi as of Q2 2026.
The equity structure is the variable that requires the most context. Databricks grants RSUs on a four-year vest schedule with a one-year cliff. Because the company remains private, those RSUs do not convert to liquid shares until a liquidity event — an IPO or a tender offer. Ghodsi told CNBC in December 2025 that he "wouldn't rule out a 2026 IPO," but more recently described the current market as not the right timing, pointing to 2027 as the more realistic window. Databricks has conducted tender offers in prior years, giving employees partial liquidity before a public event, and the pattern at pre-IPO companies of this scale suggests another tender is plausible in 2026 regardless of IPO timing. For candidates doing the math: at a $134 billion valuation, a $500,000 annual RSU grant implies a per-share price that has increased approximately 115 percent from the company's 2024 Series J valuation of $62 billion — but the liquidity date remains uncertain.
Research scientists with strong publications or documented foundation model training experience are being treated as L6 or above on arrival, reflecting the Mosaic AI team's need for senior talent with credentials that overlap with what frontier labs recruit for at the PhD entry level.
The Mosaic AI Thesis: Why Enterprise AI Talent Is Moving to Databricks
The talent migration toward Databricks is not happening in a vacuum. It is happening because of a specific combination of factors that did not exist simultaneously before 2024: open-source credibility, enterprise deployment scale, and competitive compensation.
The open-source credential matters more than most hiring analyses acknowledge. Databricks is the commercial steward of Apache Spark, the dominant distributed data processing framework, and MLflow, the de facto standard for ML experiment tracking and model registry. The $1.3 billion acquisition of MosaicML in July 2023 — which brought Naveen Rao, Jonathan Frankle, and a team of over 60 ML researchers into the company — added DBRX and the foundation model training infrastructure that Mosaic had built. That acquisition's retention packages kept the core Mosaic team intact through the critical 2023–2024 period when DBRX was being built. For ML engineers who care about open-source impact at scale, Databricks offers something OpenAI does not: public, attributable contributions to tools that tens of thousands of organizations run in production.
The enterprise deployment argument is the one Databricks recruiters lead with, and it is grounded in a real differentiator. More than 20,000 organizations worldwide run on Databricks' Lakehouse platform. An ML engineer at Databricks working on Mosaic AI's inference infrastructure is not building a research artifact — they are building the serving layer that enterprise customers will route production AI workloads through. That operational stakes dimension attracts a specific profile of ML engineer: technically rigorous, systems-oriented, and interested in seeing their work run at scale, as opposed to the research-publication profile that dominates frontier lab recruiting.
The Data + AI Summit 2026 announcements — Genie Code (a coding agent for ML engineers), AI Runtime (a unified deep learning platform for large-scale GPU training), and expanded Mosaic AI agent tooling under the Agent Bricks umbrella — signal where the next two years of engineering headcount investment will concentrate. Each of these product lines requires dedicated ML platform engineering teams, and Databricks is not in a position to staff them through internal redeployment alone at current growth rates.
The comparison to frontier labs is the one candidates and their recruiters are actively making. At OpenAI or Anthropic, a senior ML engineer is one of several thousand people working on model capability. At Databricks' Mosaic AI team — which is materially smaller — the same engineer has visible ownership of specific model behaviors or infrastructure components that ship to Fortune 500 customers. The trade-off is research-versus-deployment orientation, and for a meaningful cohort of the ML engineering market in H1 2026, the deployment path is the one they are choosing.
What's Next
Databricks has disclosed $7 billion-plus in total financing as of its February 2026 announcement — approximately $5 billion equity at $134 billion and $2 billion in debt capacity — and with annualized revenue now at $6.9 billion and AI products at $1.7 billion, the company has the financial base to sustain its hiring rate through 2026 and into a 2027 IPO window. The three product pillars accelerating that hiring — Lakebase's operational database expansion into the enterprise agent infrastructure market, Genie Code's ML engineering automation, and the Mosaic AI foundation model and inference platform — each represent open engineering headcount requirements that will not be satisfied by the current 840-role posting count alone. The most consequential signal for the US AI talent market is not that Databricks is hiring aggressively. It is that an enterprise data company has built an ML platform credible enough to pull research-caliber engineers away from the frontier labs — and is paying within range to make that case stick.
Global Career PlatformFind AI talent. Find your next role.
Booking is hotels. · Airbnb is apartments. · ENTRA is global careers.