Introduction
Databricks provides a cloud platform, which it calls a “data intelligence platform,” that lets enterprises store, process, and build machine-learning and generative AI applications on top of large volumes of data. Founded in 2013 by seven researchers from the UC Berkeley AMPLab, including CEO Ali Ghodsi and Apache Spark creator Matei Zaharia, Databricks grew directly out of open-source data-processing research before becoming one of the most highly valued private technology companies in the world.
As of mid-2026, Databricks remains private despite a valuation exceeding $130 billion, having repeatedly raised large private funding rounds rather than pursue a public listing that CEO Ali Ghodsi has said would come, but not during an unusually crowded 2026 IPO cycle.
What the company does
Databricks’ platform is built on the open-source Apache Spark data-processing engine and Delta Lake storage format that its founders originally created, combining data engineering, data warehousing, and machine-learning tooling into what the company markets as a single “lakehouse” architecture. Its Mosaic AI product line lets enterprises build, fine-tune, and deploy custom AI models and agents directly on their own governed data, positioning Databricks as an alternative to assembling separate data-engineering and AI infrastructure from multiple vendors.
Who it serves
Databricks’ customers are data engineering, data science, and machine-learning teams at large enterprises across financial services, healthcare, retail, media, and technology, among other industries. The company has said it competes directly with public companies such as Oracle and Snowflake for these accounts, and has recently expanded into adjacent categories including cybersecurity, following its move to acquire security startup Panther, and customer-data management through its CustomerLake product.
Company background
Ali Ghodsi, Matei Zaharia, Reynold Xin, Ion Stoica, Patrick Wendell, Andy Konwinski, and Arsalan Tavakoli-Shiraji founded Databricks in 2013 in San Francisco, commercializing the Apache Spark open-source project that several of them had built as researchers at UC Berkeley’s AMPLab. Ghodsi has served as CEO throughout the company’s history. Databricks has raised roughly $20 billion in total funding according to CNBC’s Disruptor 50 profile of the company, including a Series L round completed around the end of 2025 and start of 2026 that raised more than $5 billion in equity plus $2 billion in new debt capacity, at a $134 billion valuation. By June 2026, reports indicated Databricks was in talks for a further round that could value the company between $165 billion and $175 billion ahead of a potential 2027 initial public offering. Ghodsi has repeatedly said the company generates positive free cash flow and does not need to raise money to fund operations, giving it flexibility over the timing of any public listing.
Product and AI capabilities
Databricks’ Mosaic AI tools let customers train, fine-tune, and serve custom machine-learning and generative AI models using their own governed data stored on the platform, rather than relying solely on general-purpose models accessed through an external API. At its June 2026 Data and AI Summit, the company said annualized revenue had reached $6.9 billion, up more than 80% year over year, with $1.7 billion of that specifically attributed to AI products, up from $1.4 billion just a few months earlier. Ghodsi has described the growth as driven partly by AI agents themselves generating additional platform consumption — agents that query data, in turn, generate more billable usage — a dynamic he says is reshaping the company’s margin profile as customers deploy more autonomous AI workloads.
Key developments
Databricks completed a Series L funding round of more than $5 billion in equity plus $2 billion in new debt capacity at a $134 billion valuation around the turn of 2026, and reported annualized revenue exceeding $5.4 billion in February 2026, up 65% year over year. By its June 2026 Data and AI Summit, the company said annualized revenue had grown to $6.9 billion, and reports emerged the same month that Databricks was in talks for a new funding round that could value it between $165 billion and $175 billion, ahead of a possible 2027 IPO that Ghodsi has confirmed the company intends to pursue eventually.
Why it matters
Databricks’ decision to keep raising large private rounds rather than go public — even as its valuation approaches or exceeds that of long-established public software companies — illustrates how much capital is currently available to top-tier private AI infrastructure companies, and how little pressure some of them feel to test public markets during a year already absorbing enormous IPO capital from companies such as SpaceX and Anthropic. Its consumption-based pricing model, where AI agents themselves generate additional billable usage, is also a useful early example of how vendor economics shift once AI agents move from pilots into everyday production workloads.
Sector context
Within Brel’s enterprise software coverage, Databricks is the most direct competitor to Snowflake, with both companies racing to add agentic AI capabilities on top of long-established data-platform businesses. Databricks’ emphasis on custom model-building also overlaps with governed MLOps platforms such as Dataiku, though Databricks positions itself more as the underlying infrastructure layer than as a governance tool sitting on top of multiple vendors’ infrastructure.
Sources and references
This profile draws on CNBC’s reporting on Databricks’ 2026 revenue disclosures, funding rounds, and CEO commentary, cross-referenced with the company’s CNBC Disruptor 50 profile for founding and funding history.
- CNBC — “Databricks revenue growth tops 80% to $6.9 billion annualized” (2026)
- CNBC — Databricks $5 billion funding round coverage (2026)
- CNBC Disruptor 50 — Databricks profile (2026)