Most companies would prefer that you notice their technology. Unisys often succeeds when you do not. A service desk answers before a worker loses an afternoon. A bank transaction clears. A cruise ship processes another purchase. An old core system talks to a new cloud service without anyone scheduling a corporate crisis. The work is not glamorous, but the stakes are unusually clear: the machinery of a large organization must keep moving while somebody rebuilds it.
That is the useful way to understand Unisys. It is not simply a mainframe descendant, a consulting firm or a cloud reseller. It is an operator of complicated estates. Its customers have accumulated decades of applications, devices, contracts, security rules and institutional habits. They cannot throw everything away, and they cannot leave everything as it is. Unisys sells the bridge.
The company produced $1.9501 billion in revenue in 2025. Services supplied $1.611 billion of that total; software and hardware supplied $339.1 million. Its roughly 15,000 employees work across digital workplaces, cloud and application infrastructure, cybersecurity, enterprise computing and business processes. At the end of 2025, contracted backlog stood at $3.16 billion. This is a relationship business measured in renewals, migrations and years of delivery, not a download counter.
A company made from eras
Unisys Corporation dates to 1986, when Burroughs acquired Sperry and the combined business took a name assembled from “united,” “information” and “systems.” The winning suggestion beat more than 31,000 employee submissions. Its family tree reaches much farther back. E. Remington & Sons commercialized a QWERTY typewriter in 1873. William Seward Burroughs’ adding-machine business began in 1886. Later branches helped produce ENIAC and UNIVAC, then supplied computing used during the Apollo program.
Corporate history can become museum wallpaper. Here it explains the sales pitch. Unisys has spent generations translating between a new technical frontier and an institution that cannot stop operating. Today’s version of that translation is less about introducing a machine than making an unruly mix of machines behave as one system.
“We needed someone to help transform our organization.”Simon Springett, CTO, Cushman & Wakefield
Its ClearPath Forward environment is the sharpest expression of this continuity. ClearPath runs high-volume, high-availability workloads that organizations treat as core infrastructure. Unisys can maintain those systems, expose them to current development methods and connect them with cloud services. That gives the company an opening few cloud-native specialists possess: a trusted position beside the old engine room.
The product is an outcome, not a box
The portfolio falls into three reported segments. Digital Workplace Solutions manages the technology employees touch: support desks, field technicians, devices, collaboration rooms, Microsoft 365 and employee-experience analytics. Cloud, Applications & Infrastructure handles migration, modernization, FinOps, data, security and hybrid operations. Enterprise Computing Solutions contains ClearPath, application and business-process work, and specialized logistics systems.
Underneath the catalog is a simple commercial rhythm. Unisys may enter through an assessment, a migration or a workplace project. Good delivery produces knowledge of the client’s systems and trust with its operators. The relationship can then widen into a longer managed-service contract. Its annual filing calls the approach “land and expand.” The phrase is ordinary. The advantage is cumulative: every solved ticket and completed migration makes the next engagement easier to price and harder for a newcomer to displace.
The work becomes concrete in customer stories. For Omnicom, Unisys says it provides global service-desk support to more than 70,000 users, standardizes more than 80,000 devices and helped migrate eight data centers, moving more than 6,000 workloads to AWS and Azure. For Cushman & Wakefield, it reports a four-data-center move to Azure, more than 150 automated service workflows, a 73 percent reduction in major incidents and savings above a $35 million five-year goal. Across Carnival Cruise Line’s fleet, Unisys systems process more than 5.6 billion transactions annually.
The customer list also explains why the company is built globally. About 59 percent of 2025 revenue came from outside the United States. Public-sector agencies, universities, banks, manufacturers, travel operators and consumer businesses bring different regulations, but they share a preference for local delivery backed by common standards. No client supplied more than 10 percent of annual revenue. That diversity limits dependence on one account, while the concentration of certain enterprise-computing work in travel, financial services and healthcare rewards deep industry knowledge. Unisys is not merely staffing projects. It is selling familiarity with the way a regulated institution approves, changes and protects technology.
These clients are not buying the same deliverable. They are buying the same condition: fewer seams. A creative group wants hundreds of agencies to collaborate without surrendering their character. A real-estate company wants acquisitions to connect faster. A cruise line wants a transaction to work far from shore. Unisys is most relevant when technology variety has become an operating cost.
AI enters through the plumbing
Unisys’ AI strategy is deliberately applied. In 2025 it introduced Cloud AI Foundation, Cloud AI Enablement and Cloud AI Customer Experience. All three draw on the Unisys Intelligence Accelerator, a governed framework of reusable guidance and components designed to fit AI into an existing enterprise environment. One offering builds the foundation, another connects models to workflows, and the third applies conversational AI and predictive analytics to customer journeys.
That order matters. Large companies rarely lack an AI demonstration. They lack clean data, security rules, deployment discipline and a way to prove that the demonstration improved anything. Unisys’ 2026 research found a telling gap: 75 percent of surveyed leaders considered agentic AI essential to managing cloud environments, but only 23 percent had begun scaling it. The sale is not “meet the model.” It is “make the model survive procurement, governance and Monday morning.”
Most enterprise AI is not waiting for a smarter chatbot. It is waiting for the institution around the chatbot to become ready.
A June 2026 partnership with Antenna pushes the company toward measurement. The plan is to add independent benchmarks to AI-assisted software development, including visibility into delivery outcomes and token use. This is a practical response to a market moving from experimental budgets to questions about return. Unisys is betting that the next AI contract will be won with evidence, not adjectives.
Security follows the same pattern. In March 2025, Unisys launched a post-quantum cryptography service beginning with a posture assessment: inventory cryptography, identify weak points, prioritize systems and build a migration plan. Quantum computers capable of breaking current encryption are not yet commercially available. Data can still be stolen now and saved for later decryption. The first useful quantum product, in other words, may be a very thorough map.
Where Unisys fits
The competitive field is crowded. Accenture, Capgemini, Cognizant, TCS and Infosys bring enormous consulting and delivery machines. IBM and Kyndryl have deep infrastructure histories. DXC, HCLTech, NTT DATA and specialist providers compete for many of the same managed-service contracts. Cloud platforms increasingly package their own professional services, while clients build capable internal teams.
Unisys sits in a useful middle. It is smaller than the biggest consultancies and broader than a point specialist. Its differentiation is strongest when a customer needs change without interruption, particularly across regulated, international or heterogeneous environments. It can advise on a cloud move, execute it, run the resulting estate and continue caring for a proprietary core that will not disappear on schedule.
That position also creates tension. Services are labor- and trust-intensive. Large rivals can price aggressively. Clients can delay projects, bring work inside or ask automation to reduce the very support volume they once outsourced. Unisys’ answer is to automate its own delivery, move toward outcome-based pricing and sell higher-value work around AI, applications, cybersecurity and experience. It must make efficiency improve the contract rather than erase it.
The durable, unshowy promise
Unisys does not have a founder myth because modern Unisys was assembled from companies, inventions and mergers. That may be fitting. Its customers are assembled too. A contemporary enterprise is a stack of decisions made by people who never met: the database selected in 1998, the cloud contract signed in 2018, the acquisition integrated last summer, the AI pilot launched on Tuesday.
The company’s opportunity is to make that accumulated institution feel less accidental. For a buyer, Unisys can reduce support friction, modernize applications, manage cloud cost, improve security and keep high-volume systems available. For the market, it occupies the place where technological progress encounters installed reality.
A 150-year lineage does not guarantee another decade. It does offer a valuable observation: companies rarely replace one era with the next. They layer them. Unisys has made a business out of those layers, connecting the machine everyone wants to build with the one nobody can turn off.