On the afternoon of Saturday, April 18, 2009, Jonathan Schwartz — the ponytailed, blogging CEO of Sun Microsystems, a man who had inherited the wreckage of a once-great hardware company and was now presiding over its final act — picked up the phone during a Sun board meeting and dialed not
Larry Ellison, the co-founder and chief executive of Oracle, but Safra Catz, Oracle's president. Schwartz wanted more money. Oracle had offered $9.50 per share, all cash, $5.6 billion minus Sun's cash and debt — a bid that bested IBM's by a mere forty cents. Catz listened to Schwartz's counteroffer and, in the language of the subsequent securities filing, "stated would not be acceptable to Oracle." That was it. No negotiation, no counter-counter, no theatrical pause to consult with Ellison. Sun's board accepted Oracle's terms that weekend, informed IBM it was out, and by Monday morning the deal was announced. The entire trajectory of enterprise computing — the absorption of one of Silicon Valley's founding names into another — turned on a single phone call to a woman whom most of the broader business world could not have identified in a photograph.
That Schwartz called Catz and not Ellison tells you almost everything you need to know about the power structure at Oracle, and about the particular species of authority that Catz has accumulated over a quarter century. "Make no mistake," a prominent tech CEO told Fortune in 2009. "She's running the company, not Charles Phillips. When Larry and I are discussing something that goes beyond us, out comes the cellphone, and Safra is on the other end." The architecture of that arrangement — the flamboyant founder out front, the former investment banker running the machinery behind him, the extraordinary power that accrues to the person who translates vision into execution — has been the central fact of Oracle's corporate life since at least 2004, and arguably since 1999, when Catz first walked through the door.
The Numbers Behind the Ghost
By the Numbers
Safra Catz and the Oracle Empire
$920B+Oracle market cap at September 2025 peak
800%+Stock price increase during Catz's CEO tenure (2014–2025)
130+Acquisitions overseen during her tenure
$455BRemaining performance obligations (Q1 FY2026)
$3.4BCatz's estimated net worth (September 2025)
$138MTotal compensation in 2022, among highest-paid U.S. CEOs
$950KBase salary — the rest is options and stock
Holon to Brookline
Safra Ada Catz was born in December 1961 in Holon, Israel, a small city south of Tel Aviv that was founded in 1935 by thirty-two families who purchased sand dunes and built on them. Her father was a nuclear physicist; her mother, a speech therapist. When Safra was six, the family immigrated to the United States and settled in Brookline, Massachusetts — the dense, leafy suburb that wraps around Boston proper, home to JFK's birthplace and to a particular strain of intellectually ambitious immigrant families for whom education was not aspiration but oxygen.
The details of her childhood are scarce, and deliberately so. Catz has given vanishingly few personal interviews across her entire career. What emerges from the scattered biographical record is the outline of a girl whose intensity was apparent early: she was a competitive fencer at the University of Pennsylvania, where a former coach once laughingly described her as having a "combative personality." She graduated from the Wharton School in 1983 and went straight into Penn's law school, earning her Juris Doctor in 1986. She was, by the account of a Chi Omega chapter advisor who knew her as a collegian, "a very focused young woman unceasingly working towards her goals" who "approached her academic pursuits with tremendous intensity." She didn't hold a major leadership position in the sorority. She was, the advisor noted, "always there as a good, supportive member." Already, at twenty, the pattern was established: the power would come from the work, not from the title.
The law degree proved to be a credential, not a calling. In 1986, Catz joined Donaldson, Lufkin & Jenrette — DLJ, the scrappy investment bank founded in 1959 by three Harvard Business School classmates who bet that institutional research could be monetized — as an investment banker. She spent thirteen years there, rising from associate to senior vice president by 1994 and managing director by 1997. Her focus was the software industry, then in its manic adolescence, and the deals she worked on gave her an education in how technology companies were valued, acquired, integrated, and — just as important — how they destroyed themselves. "My best decision was to choose to go to Wall Street over law," she said later. "I learned a lot and focused on the expanding software industry at a time when the independent software industry was just beginning. That ultimately brought me to Oracle."
In 1997, she requested a transfer to Silicon Valley — an unusual move for a DLJ managing director, motivated, she said, by a desire to be closer to her clients. She relocated to California with her husband, Gal Tirosh, a writer, and their two young children. Two years later, in 1999, she left DLJ entirely and joined Oracle as a senior vice president. She was thirty-seven years old.
The Consigliere Arrives
To understand what Catz walked into, you have to understand Larry Ellison. Born in 1944 on the South Side of Chicago to an unwed mother who gave him up at nine months to an aunt and uncle in a modest apartment, Ellison dropped out of two universities, drove to California, and in 1977 co-founded a company called Software Development Laboratories with $2,000 — a company that became Oracle. By the time Catz arrived, Ellison was already a Silicon Valley colossus: brilliant, combative, a yacht racer and fighter-jet pilot, a man who owned a quarter of his company and whose personality was so large that it threatened to obscure the machinery that kept the enterprise running. He was also, by nature, a product visionary and a salesman, not an operator. Oracle in the late 1990s was spending freely, its margins were compressing, and the company needed someone who could impose the financial discipline that Ellison had never been interested in imposing on himself.
Catz was that person. Within months of arriving, she was named to the board of directors — in 2001 — and by 2004 she was president. The speed of her ascent was not a function of politicking or self-promotion. It was a function of demonstrated indispensability. She restructured Oracle's cost base after the tech bubble burst, driving operating profits from 21 percent of revenue in 1999 to 40 percent by fiscal year 2006. She accomplished this not through grand strategic pronouncements but through the relentless, granular work of someone who thinks in spreadsheets, not PowerPoints. "I come from Wall Street," she once said, "and you'll never see me do a PowerPoint because I'm all about Excel spreadsheets. If it's not in the numbers, I don't care how strategic it is, it doesn't play out."
The relationship between Catz and Ellison is one of those partnerships — like Buffett and Munger, or Jobs and Cook — where the chemistry between two radically different temperaments produces something neither could achieve alone. Ellison is the visionary, the showman, the man who builds cathedrals in his imagination. Catz is the engineer who figures out how to pour the foundation, source the stone, and keep the budget from ballooning. "When Larry was talking in those days about — he didn't call it the cloud, but the network and how things should actually be structured," Catz recalled, "we went on our merry way building the E-Business Suite and then we bought PeopleSoft." The use of "we" is characteristic. Catz never claims the spotlight. Her admiration for Ellison's vision is genuine and frequently expressed. But the admiration coexists with something harder: the conviction that vision without execution is decoration.
Joseph Grundfest, a Stanford securities law professor who served on Oracle's board from 2001 to 2006, put it precisely: "She's much more interested in Oracle's stock going up than she is in having her picture on the cover of a magazine."
The Acquisition Machine
If there is a single activity that defines Catz's career at Oracle, it is the deal. Over the course of her tenure, she has overseen more than 130 acquisitions — a pace and scale of dealmaking that has no real parallel in the technology industry except, perhaps, for Cisco under John Chambers. But where Chambers's acquisitions were often about buying growth and talent, Catz's acquisitions were about buying market share and then integrating the acquired companies into Oracle's operational template with a ruthlessness that terrified the industry.
The signature deal was PeopleSoft. In June 2003, Oracle launched an unsolicited — hostile — bid for PeopleSoft, the enterprise software rival, at $16 per share, valuing the company at roughly $5.1 billion. PeopleSoft's CEO, Craig Conway — a combative executive who had built his career in opposition to Ellison — publicly compared the bid to an invasion. The Department of Justice sued to block the merger on antitrust grounds. The case went to trial. Oracle won in court in September 2004, and by December, after raising its bid multiple times, completed the acquisition for $10.3 billion — more than double the original offer.
Catz was the deal's chief architect and executor. She navigated the antitrust litigation, managed the financial structure, and — critically — ran the integration after closing. That integration was savage and efficient: Oracle eliminated thousands of PeopleSoft positions, folded the product lines into its own, and extracted the cost synergies that justified the premium. The industry was horrified. Catz was unmoved. "When we do acquisitions, we decide what we want," she told Bloomberg. "We decide what fills a hole. And if the price is too high, our alternative is the $5 billion we spend on R&D every year. We're not well-known for overpaying, because at Oracle we always have an alternative."
That philosophy — the cold calculus of the alternative, the willingness to walk away, the refusal to fall in love with a deal — is pure investment banker. It is the mind of someone who has spent thirteen years on the sell side watching companies overpay for acquisitions out of ego or fear or momentum and then fail to integrate them. Catz was determined not to repeat those errors, and she didn't. After PeopleSoft came Siebel Systems ($5.85 billion in 2005), BEA Systems ($8.5 billion in 2008), Sun Microsystems ($7.4 billion in 2009), and — in her largest deal — Cerner Corporation, the healthcare IT giant, for $28.3 billion in 2022. Each acquisition was assessed, she said, "from 50 different angles." Each was integrated using what Oracle internally calls the "Oracle Playbook" — a standardized operational template, built over two decades of practice, that allows the company to absorb acquired businesses with a speed and efficiency that competitors find baffling.
When we do acquisitions, we decide what we want. We decide what fills a hole. And if the price is too high, our alternative is the $5 billion we spend on R&D every year. We're not well-known for overpaying, because at Oracle we always have an alternative.
The Playbook is not just an integration manual. It is a way of running a business. Oracle closes its financial books in fewer than ten days — faster than any other company in the S&P 500. Ninety-seven percent of its hundreds of thousands of quarterly banking transactions are automatically reconciled. The infrastructure that makes this possible is the same infrastructure that allows Oracle to swallow a $10 billion acquisition and have it operationally folded in within months. The Playbook is Catz's signature creation, and it is the reason that Oracle — despite being a 48-year-old technology company in one of the most disruption-prone industries on earth — has continued to grow its margins and its market share year after year.
The Co-CEO Experiment
In September 2014, Ellison stepped down as CEO and assumed the role of executive chairman and chief technology officer. In his place, Oracle appointed two co-CEOs: Catz and Mark Hurd. The arrangement was unusual but not unprecedented — SAP had tried it, and Deutsche Bank, with mixed results. At Oracle, it worked, because Catz and Hurd divided the company along clean operational lines: Catz oversaw operations, legal, and finance; Hurd oversaw sales, service, and marketing. They didn't step on each other because they were doing fundamentally different jobs.
Hurd was a different species of executive from Catz. A born salesman from New York City, he had run NCR and then Hewlett-Packard, where he earned a reputation as a cost-cutter of almost punitive severity before resigning under pressure in 2010 over allegations related to an expense-account scandal involving a contractor. Ellison, who had publicly defended Hurd during the HP controversy — calling his firing "the worst personnel decision since the idiots on the Apple board fired
Steve Jobs" — brought him to Oracle as co-president that same year. Hurd was gregarious, client-facing, comfortable working a room. Catz was the opposite: analytical, intensely private, allergic to the spotlight. Together, they were a complete executive.
The arrangement lasted five years. On October 18, 2019, Mark Hurd died of cancer at the age of sixty-two, just five weeks after announcing a medical leave of absence for undisclosed reasons. His death was a shock to Oracle and to the industry. Catz became sole CEO.
"After we lost Mark, all of a sudden, I was responsible for sales," Catz later reflected. "Well, what do I know about that?" The question was rhetorical — she knew plenty — but it captured something real about the transition. Catz had spent her entire Oracle career as the operational and financial brain. Now she had to be the face of the company, the one who walked into customer meetings and closed the relationship. She adapted, as she had adapted to everything else, by mastering the domain and surrounding herself with people who knew what she didn't.
The Invisible Woman
The most striking thing about Safra Catz — more striking than the deals, or the margins, or the stock price — is her invisibility. In an era when every mid-level tech founder has a personal brand, a podcast, a Twitter following, and a ghostwritten memoir, the CEO of one of the ten largest technology companies on earth has given almost no personal interviews. Oracle's own marketing department once struggled to persuade her to be photographed. She does not write op-eds. She does not cultivate journalists. She does not appear on magazine covers unless dragged there by a Fortune ranking.
This is not modesty in the conventional sense. It is strategy. Catz understood early that at Oracle, there was room for one personality, and that personality was Larry Ellison's. To compete with Ellison for attention would have been both futile and counterproductive. Instead, Catz built her power the way a consigliere does: by making herself essential, by knowing more than anyone else about the operational details of the business, by being the person the CEO calls when something needs to get done, and by never — ever — seeking credit for the outcome.
"If Catz has a personal agenda — such as becoming CEO of Oracle or another company one day — she's done an artful job of hiding it," Fortune observed in 2009. The formulation is telling: even the reporter couldn't tell whether Catz's self-effacement was genuine or performed. Perhaps it is both. The mask, worn long enough, becomes the face.
But the invisibility masks an intensity that everyone who has dealt with her describes in similar terms. The word "combative" recurs — from her fencing coach at Penn, from her negotiating counterparts, from the executives she has restructured out of existence. "You have to be better," she told an audience at the Women's High-Tech Coalition in Silicon Valley. "You have got to work harder, work longer, be louder." The advice was delivered as a statement of fact, not as a complaint. In Catz's worldview, the playing field is not level, and the correct response is not to protest the unfairness but to compensate for it with superior preparation and relentless effort.
I come from Wall Street, and you'll never see me do a PowerPoint because I'm all about Excel spreadsheets. If it's not in the numbers, I don't care how strategic it is, it doesn't play out.
Her husband, Gal Tirosh, has been credited by Catz with enabling her career by serving as a stay-at-home father to their two sons. She has acknowledged this publicly, a rare personal disclosure that reveals the architecture of trade-offs behind the career. In a culture that celebrates the work-life-balance myth, Catz has been bracingly honest about what it actually costs to operate at her level: someone has to be home.
The Cloud [Pivot](/mental-models/pivot)
For much of the 2010s, Oracle was viewed by Wall Street and the technology press as a legacy company — a database monopoly that was milking its installed base while Amazon Web Services, Microsoft Azure, and Google Cloud ate the future. The stock was flat. The narrative was death by disruption. Catz and Ellison disagreed.
Their bet was counterintuitive but, in retrospect, prescient: that large enterprises were not going to abandon Oracle's market-leading database software, and that Oracle could catch up in cloud infrastructure by doing something its competitors couldn't — offering a complete, vertically integrated stack, from database to applications to infrastructure, that could be deployed in Oracle's cloud, in a customer's own data center, or in a competitor's cloud. The multi-cloud strategy — announced through partnerships with Microsoft, Google Cloud, and AWS beginning in 2024 — was heretical. Why would Oracle let its customers run Oracle products on a competitor's infrastructure? Because, Catz reasoned, the alternative was worse: customers who felt trapped would eventually leave.
"We have always believed that the customer should get to choose what they want to use and what environment they want to run in and how to really advance their agenda," Catz explained. "Oracle doesn't believe in building walls or trapping customers." The statement was remarkable coming from a company that had spent decades extracting maximum value from its licensing agreements, a company whose sales practices were legendary in their aggression. The cultural shift was real, and it was driven from the top.
The cloud transition also required a fundamental reimagining of Oracle's business model. "The entire business model had changed from being product-based to being service-based," Catz acknowledged. The old Oracle sold licenses: build the software, throw it over the wall, collect the check. The new Oracle had to move in with the customer. "We basically get rid of the wall between our two enterprises: us and you," Catz described. Customers began telling her they couldn't "tell where my people stop and your people start" — which Catz took as the ultimate compliment.
The financial results validated the strategy. By fiscal year 2025, Oracle's cloud infrastructure revenue had reached $10.3 billion, with quarterly cloud infrastructure growth of 52 percent. The company's remaining performance obligations — a measure of contracted future revenue — stood at $455 billion as of the first quarter of fiscal year 2026, up 359 percent. Oracle was no longer a legacy company clinging to its database franchise. It was a cloud infrastructure provider being chosen by the most demanding customers in the world.
The AI Bet
And then came artificial intelligence.
The timing was almost absurdly fortunate. Just as Oracle's cloud infrastructure business was reaching scale, the explosive demand for AI training and inference capacity — driven by OpenAI, Meta, xAI, and dozens of smaller companies — created a market for exactly what Oracle had been building: dense, high-performance data centers optimized for GPU workloads. Catz moved quickly. Oracle signed cloud contracts with what she called "the who's who of AI" — OpenAI, xAI, Meta, Nvidia, AMD — and became a key participant in the Stargate initiative, a $500 billion AI infrastructure project championed by the Trump administration alongside OpenAI and SoftBank.
The numbers were staggering. In September 2025, Oracle announced a $300 billion, five-year cloud contract with OpenAI for the construction of data centers. The company projected that its cloud infrastructure revenue would reach $144 billion by fiscal year 2030 — up from $10.3 billion in fiscal year 2025. The stock surged 36 percent in a single day, the best single-day performance since 1992. Ellison's net worth briefly crossed $390 billion, making him, for a few hours, the richest person on earth. Catz's net worth jumped by approximately $412 million in the first six hours of trading, to an estimated $3.4 billion.
"Clearly, we had an amazing start to the year because Oracle has become the go-to place for AI workloads," Catz told analysts on the earnings call. The statement was characteristically understated. Oracle — the company that Wall Street had written off as a legacy database vendor — was now at the center of the most consequential technology buildout since the internet.
We have signed significant cloud contracts with the who's who of AI, including OpenAI, xAI, Meta, Nvidia, AMD, and many others.
The Healthcare Cathedral
If the AI cloud story is the most visible dimension of Catz's late-tenure strategy, the healthcare initiative may be the most ambitious. Oracle's $28.3 billion acquisition of Cerner in 2022 — the largest deal in the company's history — was not merely an enterprise software play. It was a bet that Oracle could do for healthcare what it had done for enterprise databases: build the integrated, end-to-end system that the industry needed but that nobody had been willing or able to construct.
The problem, as Catz described it, was structural fragmentation. Hospitals, clinics, and laboratories were buying software from dozens of vendors — one system for electronic health records, another for billing, another for scheduling, another for inventory — and then spending enormous resources trying to stitch it all together. "That's just not right," Catz said, "and that's coming to an end." The vision was to build a single platform — spanning clinical trials, inventory management, nursing schedules, revenue management, human capital management — that could be delivered as a cloud service and updated every ninety days.
The challenge was immense, and not primarily technical. Doctors, nurses, and radiologists were often not employees of the hospital but employees of separate groups that operated within the hospital. The billing systems were Byzantine. The data standards were inconsistent. The regulatory environment was suffocating. But Catz saw the same thing she had seen in enterprise software twenty years earlier: a market that was ripe for consolidation, integration, and the imposition of operational discipline. "We wrote that into our HCM systems," she explained. "In all of these different things — financials, billing, revenue management. It's not fair to ask hospitals to waste their resources stitching this stuff together when we do it once."
Ellison, who was driving the product vision, put it more grandly: "We're doing this because we can, but we can't do it alone." The "alone" was revealing. For a company that had historically been allergic to partnerships, Oracle's healthcare initiative required deep collaboration — with customers, with other technology vendors, with the clinical community itself. It was the most visible manifestation of the cultural transformation that Catz had been driving: from transaction optimization to customer success, from selling products to delivering outcomes.
The Political Animal
Catz's public persona is apolitical. Her actual behavior is not. In 2016, she served on
Donald Trump's presidential transition team — an unusual role for a sitting tech CEO, and one that drew criticism from Oracle employees and industry observers. She was reportedly considered for senior administration positions, including U.S. Trade Representative and National Security Advisor. She donated significantly to Republican causes and candidates. She served on the Homeland Security Advisory Council and was a member of the National Security Commission on Artificial
Intelligence.
Her relationship with Israel is deep and unapologetic. Born there, raised by Israeli immigrants, she has spoken publicly about her admiration for Israeli innovation with a fervor that goes beyond corporate diplomacy. "Israel is like an under-fire experiment," she said. "Israel doesn't follow behind the rest of the world, and it's well known Israelis don't follow anyone — even when they're asked." In a 2015 email to former Israeli Prime Minister Ehud Barak, obtained in a hack and reported by Responsible Statecraft, Catz wrote that "we have all been horrified by the growth of the BDS movement in college campuses and have concluded that we have to fight this battle before the kids even get to college. We believe that we have to embed the love and respect for Israel in the American culture." The email also mentioned that her sister, Saritz Catz, a Hollywood writer-producer, was working on a reality show called Women of the IDF, intended to "humanize the IDF in the eyes of the American public."
The disclosures caused an uproar. A group of Oracle employees published an open letter demanding that the company retract its pro-Israel stance. They cited Catz's public statements — including her reported comment that "if any employee is not with us [with Israel], we're probably not the right place for you" — and argued that the position was discriminatory and in violation of Oracle's own Code of Ethics, which affirms "the principle of equal employment opportunity without regard for any protected characteristic, including but not limited to race, religious creed, national origin, political affiliation."
Catz has not retreated. In Oracle's role as a key player in the consortium acquiring U.S. operations of TikTok — announced in September 2025 — her political and personal convictions have become newly relevant, and newly scrutinized. The intersection of her position as CEO of a company that would control a major social media platform's data and infrastructure with her documented advocacy for pro-Israel messaging has raised questions that will not be resolved by a PR statement.
The Succession
On September 22, 2025, Safra Catz stepped down as CEO of Oracle. She was sixty-three years old. The company's stock was up more than 500 percent since she had taken over as sole CEO in 2019. She had been at Oracle for twenty-six years, at or near the top of the leadership structure for twenty-one of them. Her successors were Clay Magouyrk, thirty-nine, who had been president of Oracle Cloud Infrastructure, and Mike Sicilia, fifty-four, who had been president of Oracle Industries. Catz would become executive vice chair of Oracle's board, working alongside Ellison as executive chairman.
"It is absolutely time," she told analysts. "You want to make a transition like this when things are great."
The timing was impeccable — and, depending on your interpretation, either selfless or strategic. Catz was leaving at the absolute peak: the AI tailwinds were at Oracle's back, the deal pipeline was the largest in the company's history, the stock was at an all-time high. She was not being pushed. She was not sick. She was not retreating from a scandal. She was simply — and this is the part that confounds the conventional CEO narrative — done. Or rather, not done, but reconfigured. She would remain on the board. She would remain involved in operations. She would remain, in the way that mattered most, the person Larry Ellison called when something needed to get done.
Her exit removed a female CEO from the Fortune 500. There are only eleven women running companies in the Fortune 100. The symmetry was painful and instructive: too often, as Fortune noted, women are fired under pressure after being given too little time to turn around a struggling company. Catz was the opposite case — a woman handing over a company that was soaring. If there is a glass ceiling and a glass cliff, what Catz did doesn't have a name yet.
She also teaches a course at Stanford Graduate School of Business — Mergers and Acquisitions, naturally — that she has offered nearly every academic year since at least 2018. The syllabus is not public, but the description promises "important lessons learned from the numerous M&A deals that Catz has advised on throughout her career." One imagines the students get more than lessons. One imagines they get the steel fist in the velvet glove, the same deferential grace that analysts and journalists have encountered, the quiet conviction that the numbers will tell you everything you need to know if you are disciplined enough to listen.
The Townhouse
In her keynote at Oracle's London event — one of the rare occasions when she addressed a large audience about something other than quarterly earnings — Catz used a metaphor to describe Oracle's evolved relationship with its customers. The old model, she said, was that "our geniuses in their ivory towers would build a product and throw it out to a customer. They'd lean out and say, 'Give us a call if you need help, but we're hoping you're smart enough to know how to use it.'" The new model was different. "We now have to basically move in with you. And live in the townhouse with you. And the customer's on this side and Oracle's on that side. They're totally connected, because we basically get rid of the wall between our two enterprises: us and you."
The metaphor is domestic, intimate, almost uncomfortably close. It is also, in a way Catz probably did not intend, a description of her own position within Oracle. She has lived in the townhouse with Larry Ellison for twenty-six years. The wall between her enterprise and his was eliminated long ago, if it ever existed. She has run his company's operations, executed his vision, defended his interests, expanded his empire, and enriched his fortune — which, at its peak, exceeded $390 billion. She has done this while maintaining what amounts to a state of professional invisibility, appearing only when necessary, speaking only when required, and disappearing back into the machinery the moment the lights come on.
There is a version of this story that is about self-abnegation, about a brilliant woman who subordinated her identity to a charismatic founder. But that reading misunderstands the nature of the power. Catz did not subordinate herself. She embedded herself — into the infrastructure, into the decision-making architecture, into the deal flow, into the quarterly cadence of earnings and guidance that is the heartbeat of a public company. She became, in the language of Oracle's own autonomous database, self-driving, self-securing, self-repairing. Indispensable not because of what she said but because of what she did.
On September 10, 2025, twelve days before she announced her departure, Oracle's stock surged and Catz's net worth increased by $412 million in six hours. She did not post about it on social media. She did not give an interview. She got on the earnings call and said, "It was an astonishing quarter," and then she read the numbers.
