Databricks’ AI Agent Genie Delivers Unfiltered Insight for CEOs

2–3 minutes

Multi‑step automation powered by a custom ontology

Databricks introduced an upgraded version of its AI agent Genie on Wednesday. The new release can execute complex, multi‑step workflows that previously required extended human effort. This capability stems from a proprietary ontology that Databricks built to give Genie contextual awareness of the data customers store on its platform.

CEO Ali Ghodsi puts Genie to work in real business scenarios

Chief executive Ali Ghodsi demonstrated Genie’s internal use, moving beyond a typical product demo to share actual chat logs and analytics. He showed how the agent dissected an employee satisfaction survey, highlighted the areas where the company scored poorly, and helped evaluate and discard proposed fixes that executives had suggested.

In another example, Genie summarized a problematic sales pitch, flagging the features the prospect demanded before agreeing to buy. The agent also prepared Ghodsi for a podcast interview by reviewing his past appearances and drafting likely questions. While we spoke, Genie even suggested what he should say next, offering surprisingly sound advice.

Seeking the truth that a CEO no longer hears

The most striking aspect of the demonstration was Ghodsi’s reliance on Genie to surface information he would otherwise miss as the head of a 12,000‑person organization. The agent served as a source of unvarnished feedback about problems that needed attention.

“Have you heard of Dunbar’s number? The magic number is 150,” Ghodsi said when I pointed this out. Named for British anthropologist Robin Dunbar, it’s theoretically the maximum number of people with whom humans can maintain stable relationships. In a small startup, a CEO knows everyone and can receive candid negative feedback directly.

With a workforce far larger than 150, Ghodsi acknowledges he cannot maintain personal connections with every employee. Consequently, a salesperson is unlikely to admit a disastrous call to the CEO, and an HR manager may not confess that a proposed employee‑satisfaction initiative is doomed to fail.

“Anyone who has more than 150 employees, which is a lot of companies, a lot of people can get value out of this,” he said, referring broadly to AI agents and specifically to Genie.

From dogfooding to a product for business leaders

Ghodsi’s practice of using Genie internally—often called “dogfooding”—has become a cornerstone of Databricks’ pitch for the agent as an always‑on tool for executives. He also uses Genie in his personal life, opting for it over security‑questionable alternatives such as OpenClaw, Instinct, or Muse.

He envisions Genie delivering to business leaders the same productivity gains that AI agents and harnesses have delivered to programmers. Regarding concerns that agents could drive up corporate AI costs, Ghodsi points to Databricks’ Unity Gateway. The product lets organizations set usage limits and substitute cheaper open‑weight models when needed.