Databricks has crossed a $7 billion revenue run rate, growing 80% year over year, and agents now send it more queries than people do, investor Ben Horowitz wrote in an essay, “Databricks & the Future of the Firm,” published on X on 27 October 2026.
Horowitz writes that his firm, a16z, led Databricks’ Series A in 2013. He uses the company to ask what people will do at work all day once agents arrive, and writes: “It might be closer to, “Everyone is a manager.””
Short explainer · Horowitz on Databricks
A vertical news Short narrated from the published text of this piece and related FourWeekMBA pieces.
An evergreen Business Engineer explainer on why products built on the same model differ. It explains the general idea and does not discuss Databricks.
The key insight: As we read it, Horowitz’s essay puts the agent story in database terms: by his account, agents now send Databricks more queries than people, and the firm’s system of record decides what those agents can know and do.
Horowitz writes that last month Databricks “crossed a $7 billion revenue run rate, growing 80% year over year, and was valued at $190 billion.” Databricks is private, and the essay is the source for these figures.
He breaks out two products. Databricks’ Lakehouse “runs at a more than $1.5 billion revenue run-rate, growing over 100% year over year,” he writes. Lakebase, the operational database built on the Neon acquisition, “launched last year and has already crossed a $100 million revenue run rate, and agents are its most prolific users.”
The essay opens with the agent figures. “Three years ago, every interaction on Databricks involved a human,” Horowitz writes. “Today, the company is seeing more queries from agents than people, and more than 17 million databases are launched by agents every day.”
Revenue run rates as given in Ben Horowitz’s essay of 27 October 2026: Databricks at more than $7 billion, growing 80% year over year; its Lakehouse at more than $1.5 billion, growing over 100%; and Lakebase at more than $100 million. The essay gives no further breakdown. Databricks is private and these are not audited figures.
Ben Horowitz published the essay as an X article on 27 October 2026. It is the source of every figure and quote in this piece, including the $7 billion run rate, the 17 million databases a day and Ali Ghodsi’s ‘system of record for agents’.
A System of Record for Agents
Two figures from Ben Horowitz’s essay of 27 October 2026: Databricks’ revenue run rate of more than $7 billion, growing 80% year over year, and more than 17 million databases launched by agents every day. Databricks is private; these are not audited figures.
The essay quotes Databricks CEO Ali Ghodsi on why agents need a new kind of record. “AI does not have an intelligence problem,” Ghodsi says. “It has a context problem.”
Ghodsi’s answer is the lakehouse. “The lakehouse has become the system of record nowadays,” he says in the essay. “I call it SORA – the System of Record for Agents.”
Horowitz writes that the record has to answer four questions: what is true, what it means, what is allowed and whether it worked. On meaning, he reports that in Databricks’ testing, grounding an agent in the company’s ontology “raised the accuracy of answers from about 50% to 84.5%.”
Everyone Is a Manager
Horowitz frames the shift as a second managerial revolution, after James Burnham’s book The Managerial Revolution. He writes that “the decline of managerialism may look more like everyone becoming a manager.”
“The second managerial revolution is not so much the death of management as it is the end of the professional managerial class’s monopoly on it,” he writes.
He points to software engineers first. “An engineer’s day increasingly looks like a manager’s: scope the problem, break it into tasks, delegate them to agents working in parallel, then review the pull requests that come back,” he writes.
The essay also sets Ghodsi against Block’s Jack Dorsey, who wrote in March about replacing hierarchy with a company “world model.” Ghodsi is quoted saying, “I agree philosophically with Jack Dorsey,” before asking, “The question is does that apply to the org chart of humans too?”
The Sovereign Firm
Horowitz’s case for the firm owning its data rests on two banks renting the same frontier model. Only the one that records the facts behind each answer, human overrides and outcomes builds an advantage, he argues.
“If you replace the AI model at your company and the new one inherits your definitions, permissions, evaluated cases, customer history, and hard-won exceptions, that is a foundation for a sustainable competitive advantage,” he writes.
As we read it, the numbers come from an interested party: Horowitz writes that a16z led Databricks’ Series A and, with NEA, its Series C, and the essay does not cite a filing.
As we read it, the essay’s claim about work covers both roles and headcount: Horowitz writes that more people will be specifying work, delegating execution and inspecting results, and that ten people will be able to do what used to take thousands.
As we read it, the competitive argument is about memory: Horowitz writes that a firm that keeps its definitions, permissions and evaluated cases can swap models without starting from the same prompt as everybody else.
Ali Ghodsi, Databricks CEO, as quoted by Ben Horowitz, 27 October 2026
“AI does not have an intelligence problem”
Three Implications
FOR SOFTWARE BUYERS Ghodsi says in the essay that the lakehouse has become the system of record, which puts Databricks in the fight Horowitz describes with earlier systems of record.
FOR MANAGERS Horowitz writes that the decline of managerialism may look more like everyone becoming a manager.
FOR ENGINEERS Horowitz writes that an engineer’s day increasingly looks like a manager’s, delegating tasks to agents and reviewing what comes back.
The framework puts it this way: “Enterprise vendors don’t sell software anymore—they sell irreversibility. The product is the interconnection.”
As we read it, Horowitz’s essay makes a version of that case for the agent era: Ghodsi calls the lakehouse the system of record for agents, and Horowitz argues the firm’s advantage is a record that outlasts any one model. The framework was written in December 2025 and does not discuss Databricks.
What Is Not Established
The revenue, valuation and agent figures come from Horowitz’s essay, which does not cite a filing. Databricks is a private company, so they are not audited public figures.
Horowitz writes that a16z led Databricks’ Series A and, with NEA, its Series C, so he is an investor describing a portfolio company. The 84.5% accuracy figure is from Databricks’ own testing, as the essay states.
This Week In The AI Supercycle
Execution Fault Lines, Concentrated Demand, and Geopolitical Firewalls
This week’s issue: the AI build is not short of demand for its parts. It is short of cheap money, and its end revenue rests on a narrow base of large, often venture-funded buyers.
Ben Horowitz writes that Databricks has passed a $7 billion run rate with agents now its biggest source of queries, and he writes that more workers will specify, delegate and judge work done by agents.
A note on sourcing. All figures and quotes are from Ben Horowitz’s X article “Databricks & the Future of the Firm,” published on 27 October 2026, which we read in full; quotes of Ali Ghodsi and Jack Dorsey are as given in the essay. Horowitz’s firm a16z is a Databricks investor, as the essay states. This is not investment advice.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
