Unisys

unisys.com

Unisys combines project-based technology delivery, continuing IT operations and enterprise-computing products rather than operating solely as an advisory firm or software vendor. Buyers can engage it for application engineering, enterprise-system implementation, cloud migration, cybersecurity and employee support. Its separately documented AI services cover readiness and roadmaps, solution development, and Copilot implementation and operation. Procurement should distinguish advisory deliverables, implementation responsibilities, managed-service commitments and product licensing, particularly where an engagement involves other delivery partners. ^[1] ^[2] ^[3] ^[4] ^[5] ^[6] ^[7] ^[8]

Services and delivery

AI readiness and roadmap advisory

Advisory. Assesses infrastructure, data quality and skills readiness; identifies AI use cases; and develops adoption roadmaps. Deliverables include executive workshops and advice on secure, ethical adoption, rather than an assumed production deployment. ^[1]

AI solution design and development

Implementation. Designs and develops AI solutions with supporting architecture and change management. The offering also describes performance measurement and optimization; buyers should establish the exact production and model-operation responsibilities. ^[2]

Microsoft Copilot deployment

Implementation. Assesses the environment, business case, user personas and licensing strategy, then prepares Microsoft 365 and Teams, addresses data security, educates users and customizes business-workflow integrations. ^[9]

Microsoft Copilot managed services

Managed services. Provides continuing Copilot support, maintenance, monitoring and optimization, including security and compliance work. This is an operational service around the product, not simply a license sale. ^[9]

Application modernization and digital product engineering

Implementation. Modernizes applications and delivers digital products through design, development, testing and deployment. This is software implementation; use of AI-enabled tooling does not make every engagement a custom AI project. ^[3]

Enterprise application implementation

Implementation. Implements enterprise applications, with SAP, Oracle, Salesforce and Pega named in the offering. Buyers should confirm the relevant product modules and the division of integration and migration responsibilities. ^[3]

Managed application operations

Managed services. Provides continuing management of enterprise applications, distinct from initial development or modernization projects. ^[3]

Software quality engineering

Implementation. Delivers automated checks, regression testing, security testing and compliance reporting to support application quality and release assurance. ^[3]

Hybrid-cloud transformation

Implementation. Assesses, migrates and modernizes applications, infrastructure and workloads across public, private and hybrid-cloud environments. ^[4]

Hybrid-cloud managed services

Managed services. Operates cloud environments through automation, observability, governance, performance management, security and cost optimization. ^[4]

Cybersecurity assessment and governance

Advisory. Provides security assessments and governance advice to identify exposure and define security requirements before implementation. ^[5]

Security architecture and implementation

Implementation. Designs and implements controls including identity and access management, network segmentation and cyber recovery. Scope should identify which controls Unisys implements and which remain customer responsibilities. ^[5]

Managed security operations

Managed services. Offers managed detection and response with round-the-clock monitoring and incident management, alongside attack-surface and vulnerability management, managed SASE and SD-WAN, and managed cyber recovery. ^[5]

Data transformation, governance and analytics

Implementation. Delivers data discovery, profiling, quality assessment, migration and transformation, including preparation for data lakes, visualization and self-service analytics. These services do not independently establish AI development. ^[10]

Managed employee IT support

Managed services. Provides workplace IT support through its Agentic Service Desk, using Service Experience Accelerator technology and agentic automation. This is a support service, not evidence of bespoke AI-model engineering. ^[6]

ClearPath computing products

Software product. Supplies ClearPath Libra and Dorado systems and MCP and OS 2200 operating environments. Product provision is separate from modernization or managed-service engagements. ^[7]

ClearPath modernization and cloud transition

Implementation. Provides application modernization and AWS or Azure transition services for ClearPath environments, separate from purchasing computing products. ^[7]

ClearPath managed services

Managed services. Provides ongoing managed services for ClearPath environments. Buyers should specify the operating environment, application boundaries and service levels covered. ^[7]

Company details

Track record

For NI, Unisys reports developing a generative-AI modernization assistant in nine weeks using OpenAI’s Assistant and GPT-4o models, packaged as an Eclipse plugin. Functions included complexity assessment, Java and microservices modernization, JUnit generation and API documentation. The case describes CI/CD integration and manual review, but says exact outcome measurements were still being gathered. ^[11]

At the Austrian Ministry of Finance, Unisys reports deploying a Smart Rule pilot to extract business rules from legal text concerning study grants and family funding, with legal-expert review. Automating incoming citizen requests through a rule-based engine was explicitly outside the pilot. ^[12]

For Benjamin Moore, Unisys reports migrating approximately 200 non-SAP servers to Azure, implementing a landing zone and ExpressRoute connectivity, and deploying MDR/SIEM services. Partner oXya handled approximately 100 SAP servers and the RISE with SAP transition. Unisys reports a 19% Azure cost reduction over 12 months for the non-SAP workloads; this is a project-specific result. ^[8]

Scope and limitations

The supplied evidence is official and self-reported; it does not independently validate delivery quality or outcomes.
The Copilot service sheet has a January 2024 version marker. Current scope, licensing prerequisites and commercial terms require confirmation.
The NI narrative retains a Q2 2025 completion target. It does not confirm present project status or establish whether repository adaptation involved fine-tuning, retrieval or another method.
AI delivery evidence supports solution development and specific client implementations, not every machine-learning discipline. Detailed model-operation responsibilities remain a contractual question.
The Austrian implementation was a pilot, while the separate human-centered AI story concerns Unisys’s internal workflows rather than an external customer reference.

Questions to ask

Which work will Unisys deliver directly, and which responsibilities belong to partners, subcontractors or our team?
Can you provide current production references matching our proposed applications, platforms and operating requirements?
For AI work, what model choices, data-handling controls, evaluation methods and acceptance criteria will apply?
Who owns custom assistants, plugins, integrations and generated code, and how are third-party licensing obligations handled?
What support coverage, incident-response commitments and model-change responsibilities are included after deployment?
What are the licensing prerequisites, transition costs, service levels, exit arrangements and handover deliverables?

Overview