What is Databricks?

What is Databricks?

Databricks is the company behind the Data Intelligence Platform, a cloud-based system that unifies data engineering, analytics, data warehousing, and AI/ML on a single 'lakehouse' architecture. Founded in 2013 by the UC Berkeley team that created Apache Spark, it now serves more than 20,000 organizations and, as of its June 2026 Data + AI Summit, runs at a $6.9 billion revenue run-rate growing more than 80% year-over-year.

Databricks pioneered the 'lakehouse' — an architecture that merges the cheap, open storage of a data lake with the governance and SQL performance of a data warehouse, built on open-source projects it created or stewards: Apache Spark, Delta Lake, MLflow, and Unity Catalog. The platform runs natively on all three major clouds (AWS, Azure, and Google Cloud) and lets customers run ETL, BI/SQL, machine learning, and generative-AI workloads against one copy of their data.

The scale is large. At its June 16, 2026 Data + AI Summit, Databricks said it had crossed a $6.9 billion annualized revenue run-rate growing more than 80% year-over-year — a sharp acceleration from the $5.4 billion (+65%) it reported in February 2026. It serves more than 20,000 customers, over 800 of them spending more than $1 million a year and more than 70 spending over $10 million, with adoption by more than 60% of the Fortune 500. Net revenue retention exceeds 140%, and the company has been free-cash-flow positive over the trailing twelve months.

AI is now the fastest-growing part of the business: Databricks' AI products (its Mosaic AI stack, Genie assistant, and agent tooling) reached a $1.7 billion annualized run-rate, up from $1.4 billion four months earlier. That growth carries a cost — gross margins have slipped to around 74% (from above 80%) as AI agents drive heavier compute consumption. Gartner has named Databricks a Leader in its Magic Quadrant for Data Science and Machine Learning Platforms, positioning it as the primary challenger to Snowflake and the cloud hyperscalers' native analytics stacks.

Sources:CNBC — Databricks revenue tops 80% to $6.9B annualized (Jun 2026)Databricks newsroom — $5.4B run-rate, Lakebase & Genie (Feb 2026)Databricks — Data Lakehouse product

What does Databricks offer?

A unified Data Intelligence Platform spanning data engineering, warehousing, governance, and AI — built on open lakehouse foundations.

Sources:Databricks — Data Lakehouse productDatabricks newsroom — Lakebase & Genie (Feb 2026)

How does Databricks make money?

Databricks sells consumption-based access to its platform, billed in Databricks Units (DBUs) on top of the customer's cloud spend, plus subscription tiers that unlock governance, security, and AI features. Revenue is driven by usage growth inside existing accounts (net retention >140%) far more than by per-seat licensing.

The core unit of pricing is the DBU — a normalized measure of compute consumed per second. Rates vary by workload and tier: SQL Classic compute runs around $0.22/DBU on AWS, all-purpose compute is roughly $0.55/DBU on the Premium tier, and fully managed Serverless SQL is about $0.70/DBU in US regions (higher in the EU). Customers also pay their cloud provider separately for the underlying VMs and storage, except on serverless, where infrastructure is bundled into the DBU rate.

Databricks packages features into tiers — historically Standard, Premium, and Enterprise. Premium (now effectively the default, including Unity Catalog, role-based access control, SQL warehouses, serverless, and the full Mosaic AI suite) is the volume tier; Enterprise adds compliance certifications, dedicated support, and custom SLAs at roughly 15–25% higher DBU rates. The legacy Standard tier is being retired across clouds, pushing accounts up to Premium.

Growth comes from landing a workload and expanding consumption as more teams move data engineering, BI, and AI onto the platform — which is why net revenue retention exceeds 140%. Large enterprises sign committed-use contracts (DBCUs) for volume discounts, and the AI products plus the newer Lakebase serverless database are the fastest-growing consumption drivers — AI alone reached a $1.7 billion run-rate, and the company's data-warehousing business also crossed $1 billion. The flip side of agent-heavy consumption is margin: gross margins have eased to about 74% as AI workloads burn more compute.

Sources:Flexera — Databricks pricing guide (DBU tiers)CNBC — Databricks $6.9B run-rate, AI at $1.7B, ~74% margins

Who leads Databricks?

Databricks is led by co-founder and CEO Ali Ghodsi, alongside the other UC Berkeley 'Apache Spark' founders and a public-company-caliber executive bench (CFO Dave Conte, plus revenue and operations leadership).

Sources:Databricks — Leadership teamDatabricks — Founders

How do you contact Databricks's leadership?

Databricks does not publish individual executive email addresses, but its verified corporate pattern is first.last@databricks.com (used in ~86% of addresses). The addresses below follow that verified format and are inferred, not officially published; for press, Databricks routes through press@databricks.com and its newsroom.

• Ali GhodsiCo-founder & CEOali.ghodsi@databricks.com
• Dave ConteChief Financial Officerdave.conte@databricks.com
• Ron GabriskoChief Revenue Officerron.gabrisko@databricks.com
• Press / MediaMedia relationspress@databricks.com

Sources:RocketReach — Databricks email format (first.last, 86%)Databricks — Newsroom / press

How much funding has Databricks raised?

Databricks has raised roughly $19 billion in equity across twelve disclosed rounds (Series A through L), most recently a Series L of more than $4 billion at a $134 billion valuation announced in December 2025 — making it one of the most valuable venture-backed companies in the world.

The early rounds were Andreessen Horowitz- and NEA-led: Series A in September 2013 ($13.9M, led by a16z), Series B in 2014 ($33M, NEA), Series C in 2016 ($60M, NEA), and Series D in 2017 ($140M, a16z). Series E in February 2019 raised $250M at a $2.75B valuation (a16z), followed by Series F in October 2019 ($400M at $6.2B, a16z).

The scale-up rounds came fast: Series G in February 2021 raised $1B at a $28B valuation (led by Franklin Templeton), Series H in August 2021 raised $1.6B at $38B (led by Morgan Stanley's Counterpoint), and Series I in September 2023 added $500M at a $43B valuation (co-led by Capital One Ventures and Nvidia). Series J in December 2024 was a landmark $10B raise at a $62B valuation led by Thrive Capital (co-led by a16z, DST Global, GIC, Insight Partners, and WCM).

The most recent rounds reflect the AI-driven re-rating. Series K in September 2025 raised about $1B at a $100B-plus valuation (co-led by Thrive Capital, a16z, Insight Partners, MGX, and WCM). Then Series L, announced December 16, 2025, brought in more than $4 billion of equity (about $5B) at a $134B valuation — led by Insight Partners, Fidelity, and J.P. Morgan Asset Management — alongside roughly $2B of new debt, for over $7B of total capital. Databricks is free-cash-flow positive, and CEO Ali Ghodsi has signaled it is IPO-bound, with an S-1 widely expected around Q3 2026 and a debut in late 2026 or 2027.

Sources:Built In — Databricks $4B+ Series L at $134B valuationDatabricks newsroom — $10B Series J at $62B valuationWikipedia — Databricks funding history

How did Databricks get here?

From an open-source Spark project out of UC Berkeley to a $134B data-and-AI platform running at a $6.9B revenue run-rate in roughly a decade.

Sources:CNBC — Databricks revenue tops 80% to $6.9B annualizedTechCrunch — Databricks to buy Neon for $1BWikipedia — Databricks history & acquisitions

Who are Databricks's competitors?

Databricks competes with the cloud data warehouses and the hyperscalers' native analytics stacks, with Snowflake as its closest head-to-head rival.

Sources:CNBC — Databricks worth more than SnowflakeFlexera — Databricks vs Snowflake