A research paper gave Larry Ellison a map
Larry Ellison found the opening in a research paper. In the IBM Journal of Research and Development, he came across a description of a working prototype for a relational database management system, an idea that suggested a practical route into a new kind of software, according to Oracle’s corporate timeline. The paper mattered because it turned a large technical question into a direction: data could be arranged through relationships, queried systematically, and made useful to the organizations that depended on it.
That direction began with the relational model outlined by British-born computer scientist Edgar F. Codd. Codd’s work supplied the conceptual frame that inspired Ellison and his colleagues, as Britannica recounts. The model offered a disciplined way to think about information: records could sit in related structures, and the value of one piece of data could emerge through its connection to another. For people building software, that idea pointed toward a product with a clear purpose, even while the market around it was still forming.
Ellison’s discovery therefore carried a particular kind of risk and promise. A research prototype showed that the approach could work; it left open the harder commercial questions of how to turn that approach into software people would buy, run, and depend on. The opportunity sat in the gap between a theory of organizing data and a product capable of defining a category.
That gap gives Oracle’s origin story its force. The company’s future began with a technical argument on paper, then moved toward the demanding work of making the argument real. Before a database could become a business, the idea needed builders, a name, and a place to start.
A CIA Project Supplied the Name and the Opening
The builders arrived with a working history of their own. Larry Ellison, Bob Miner, and Ed Oates formed Software Development Laboratories in 1977 after their time together at Ampex, where Ellison and Miner worked as computer programmers and Oates served as Ellison’s supervisor, according to Britannica. Their roles gave the new company a compact founding team: two engineers and the manager who had already seen Ellison at work. That shared history mattered because the task ahead demanded more than interest in a promising technical model. It demanded people who could turn an idea into software, organize the work around it, and keep pushing until customers had something they could use.
An early CIA project gave that effort its name. Ellison, Miner, and Oates had worked on a project called Oracle near the beginning of their association, and the database inherited the name, according to Oracle’s corporate timeline. The word carried a useful sense of purpose for a database: a place organizations might turn to ask questions of their information and receive an answer. It also gave the young group a label that could travel easily, from an earlier assignment into a new company and eventually onto a product. The technical argument from the paper now had a human frame around it, shaped by three people whose connection had begun in a workplace and continued through a government project.
They began on a scale that made every choice visible. Software Development Laboratories’ first office measured 900 square feet in Santa Clara, California, as Oracle describes its history. Inside that small space, the founders had the name, the team, and a clear reason to build. The office also set the practical terms of the venture: a company small enough for its early work to depend directly on the people in the room, yet pointed toward software meant to serve organizations with far larger stores of information. Oracle was still a name attached to an ambition. The next step was to make that name belong to a real database.
The first Oracle had to fit inside 128K
The company was small enough for its early work to depend directly on the limits of the machine it would have to serve. Oracle Version 1 ran on a PDP-11 under RSX, had been written in assembly language, and used 128K of memory, according to Oracle’s corporate timeline. Those details place the founders’ ambition in a sharply defined working environment: every part of the early database had to earn its place inside a small technical envelope, on a specific machine and operating system.
That constraint gave the project a practical shape. A relational database promised a new way to handle information, yet the first task was intensely concrete: make the software run within the memory available, in the language closest to the machine. Assembly language put the work directly alongside the PDP-11 and RSX environment, while 128K gave the team a hard boundary around what the first version could hold. The database began as an engineering exercise with very little room for excess, and that narrow space forced clarity about what the product had to do.
Oracle Version 1 remained an internal milestone rather than a public release; it was never officially released, as Oracle records in its timeline. The next version carried the idea into the market. In 1979, Oracle Version 2 arrived as the first commercial SQL relational database management system, according to Oracle. The jump from Version 1 to Version 2 matters because it marks the moment when the founders’ early technical work became software a customer could actually buy and use.
That customer arrived through the U.S. Air Force, which became Oracle’s first customer and used the program at Wright-Patterson Air Force Base near Dayton, Ohio, as Britannica reports. A database that had started within 128K of memory had reached an operating setting where its value could be tested against real institutional work. The product now had its first proof of use, and the company had a foundation for the much larger opportunity ahead.
Portability turned a database into a business
That customer arrived through the U.S. Air Force, which became Oracle’s first customer and used the program at Wright-Patterson Air Force Base near Dayton, Ohio, as Britannica reports. A database that had started within 128K of memory had reached an operating setting where its value could be tested against real institutional work. The product now had its first proof of use, and the company had a larger technical challenge in front of it: carrying that product into the many computing environments where potential customers already worked.
Oracle Version 3 answered that challenge through C. The version was written in the language and ran on hundreds of operating systems and hardware platforms, according to Oracle’s corporate timeline. That reach changed the practical terms of the business. A customer could encounter Oracle as software suited to its own machine rather than as a product tied to one particular setup, and every additional platform created another path for the database to enter an organization.
SQL gave those paths a common language. The database had commercial roots in SQL, while the language itself gained formal reference points through the American standard ANSI X3.135-1986 in 1986 and the international standard ISO 9075:1987, published in January 1988, as documented in IEEE Annals of the History of Computing. For Oracle, that developing standard mattered on both sides of the company. Engineers could build around a shared database language, while sales teams could present customers with a product connected to a growing technical convention. Portability and SQL therefore worked together: one widened the range of machines Oracle could serve, and the other supplied a recognizable way to work with data across those machines.
Andrew Mendelsohn, who worked on the software development team that built the Oracle relational database management system, later traced the company’s rise from a small niche software business to one of the industry’s giants across the mainframe, minicomputer, client-server, and Internet computing eras. His account places the source of that rise in a highly innovative RDBMS that stood apart from competitors, joined by strong marketing and sales organizations, according to the IEEE history. Version 3 helps make that assessment concrete. The technical decision to run widely gave the sales organization a broader field of customers to pursue, and the commercial organization gave the technical advantage a way to travel. Oracle’s database could now move with the changing computer landscape, carrying the company toward a much larger stage.
Oracle scaled at the speed of the 1980s software boom
That commercial force soon had to carry a company whose identity was changing as quickly as its market. Software Development Laboratories became Relational Software Inc. in 1979, then entered its Oracle Systems Corporation era in 1982, when the flagship product supplied the name that would endure, according to Britannica. Each change marked a sharper public expression of the same ambition: sell a database system widely enough that the company could grow with the computing industry around it.
The public markets offered the next mechanism for that growth. On March 15, 1986, Oracle sold 2.1 million shares in an initial public offering on NASDAQ; the company then had 450 employees and US$55 million in annual revenue, according to Oracle’s corporate timeline. Oracle entered the market during a technology boom that also brought public offerings from Microsoft, Sun Microsystems, and Adobe Systems. The timing gave the company a place in a widening class of software businesses whose products had become central to the organizations buying computers in ever greater numbers.
Oracle moved rapidly from participant to category leader. In 1987, it ranked as the world’s largest database management company, reporting US$100 million in sales and 4,500 end users across 55 countries, according to Oracle. Those figures give the earlier sales and marketing strength a practical shape: a product built for many platforms had found customers across national borders, and the company had learned how to turn that reach into a growing installed base.
By the end of the decade, Oracle’s annual revenue had risen above US$500 million, while its corporate campus stood in Redwood Shores, California, as Oracle’s timeline records. The company also joined the S&P 500 in 1989 and moved to Redwood Shores that year, according to Oracle. The shift in scale changed the stakes for the database maker: a fast-growing public company now had the resources, the customer base, and the institutional presence to carry its software further into the operations of large enterprises.
The Product Moved Deeper Into the Enterprise
Oracle used those resources to push its software further into the daily machinery of large organizations. Oracle7 arrived in 1992, extending the company’s database line at a moment when its customers were building ever larger systems around it, according to Oracle. Three years later, Larry Ellison introduced a strategy for delivering Oracle software via the internet, placing the company’s next move in a networked future rather than within the boundaries of a single machine or office, as Oracle records. The direction mattered because databases had already become central repositories for corporate information, and Oracle was now developing the means to place related software closer to the work that information supported.
That broader role took shape through applications as well as the database itself. In 1999, InformationWeek named Oracle Database one of the most influential products of the 1990s, a recognition that captured the product’s growing place in business computing, according to Oracle. The company then put its own enterprise-software ambitions through a demanding internal test. Oracle says it saved US$1 billion in 2000 by consolidating its systems with Oracle E-Business Suite 11i, using the suite to bring its operations together on its own technology, according to the company.
Oracle described E-Business Suite 11i as the first fully integrated enterprise apps suite, and that description points to the shift underway inside the company’s portfolio, according to Oracle. A database held the records. Applications organized the work around those records. Operational systems carried that work across finance, staffing, customers, and the other routines that keep a large institution moving. As those layers came together, Oracle’s software could sit nearer the decisions and processes that shaped an organization’s day. The company’s influence therefore rested increasingly on the connections between systems, where a change in one part of a business could travel through many others.
That position also gave Oracle a clear reason to expand its reach through the products companies already used. In 2005, Oracle acquired PeopleSoft, bringing the story to the threshold of an era in which acquisitions would become a major way to assemble a wider enterprise technology portfolio, according to Oracle.
Acquisitions and cloud kept extending Oracle’s reach
That wider portfolio became a central engine of Oracle’s growth, as the company pursued software businesses serving a range of business and technology applications. Britannica describes much of Oracle’s expansion as the result of aggressive acquisitions, a strategy that gave the company new products and brought its systems into more parts of a customer’s operations. The pattern turned each deal into a practical question of reach: which additional layer of enterprise technology could Oracle carry alongside its database and applications?
Its 2010 acquisition of Sun Microsystems supplied a defining answer. Through Sun, Oracle acquired Solaris and became the steward of Java, according to Oracle, two technologies with their own deep places in the software landscape. The deal therefore widened the company’s responsibilities as well as its product range, connecting Oracle to technologies used to build, run and support software across organizations. That larger position arrived as enterprise computing was shifting toward cloud delivery, and Oracle moved its own product line in that direction.
Oracle rolled out Fusion Cloud Applications in 2011, then introduced multitenant architecture for cloud in Oracle Database in 2013. Oracle’s corporate history presents those releases as steps in a continuing effort to adapt its established enterprise software for a new operating model. The multitenant design placed the database directly within that cloud transition, while Fusion Cloud Applications extended the company’s application portfolio into the same era. Together, they made cloud a new setting for the reach Oracle had spent decades building through databases, business applications and acquisitions.
In 2018, Oracle launched Autonomous Database, which it calls the industry’s first database to be self-patching, self-tuning and self-managing. According to Oracle, the product carried the company’s database work into a model designed around automation, adding another way for customers to run core systems. The move followed the logic of the broader portfolio: Oracle kept extending what it could provide around the data and applications that sit at the center of enterprise work.
The scale of that acquisition-led strategy now appears in Oracle’s own tally of more than US$110 billion spent on more than 150 acquisitions. The company reports that figure alongside its continuing investment in research and development, showing how purchases and product development have operated together. By this point in the story, Oracle had built a far wider technology estate than the early database company, setting up the question of what that accumulated reach looks like at full corporate scale.
A 128K experiment became enterprise infrastructure
Oracle had built a far wider technology estate than the early database that first gave the company its name, and the change shows most clearly in the scale of the organization carrying that estate forward. Twenty years after its 1986 public offering, Oracle had grown to a global workforce of 65,000 and annual revenue above US$15 billion, according to Oracle’s corporate timeline. The following year brought a similarly revealing snapshot: Oracle reported US$18 billion in revenue, 65,000 employees and 275,000 customers in more than 145 countries in 2007, figures that describe a company woven into the operations of organizations across a broad international footprint, according to Oracle.
That footprint has continued to expand in both revenue and the range of people required to serve it. Oracle reports US$57 billion in FY2025 revenue and 141,000 employees, a workforce that includes 29,000 consulting experts and 18,000 customer support and service specialists speaking more than 20 languages, according to the company. Those roles give the numbers their practical meaning: software built around data now travels with implementation work, ongoing service and expertise delivered across languages and regions.
Oracle expresses the purpose of that work in direct terms: to help people see data in new ways, discover insights and unlock endless possibilities, in the company’s own words. The promise carries the original commercial insight into its present scale. Data became the product, then the platform, and finally a durable point of leverage for the organizations that depend on it.
Comments (0)