NelsonHall recently completed an in-depth analysis of cognitive & self-healing IT infrastructure management services, researching the capabilities of leading IT services vendors and the requirements of their clients. This blog looks at the investments vendors need to make to meet client demand and how the market will evolve over the next 12 to 18 months.

Enterprises utilize cognitive IT infrastructure services to increase predictability across the IT ecosystem, enhance experience, and improve business outcomes. AI now has a huge and increasing role to play here, resulting in:

• Expanding investment in AI (AIOps, GenAI, and Agentic AI) to support cognitive IT infrastructure
• Increasing demand for SRE-led operations to improve reliability and experience
• Greater focus on OCM and AI strategy and consulting services to drive technology adoption and ROI.

Expanding investment in AIOps, GenAI & Agentic AI to support cognitive IT infrastructure

The investment in AIOps continues in support of monitoring, self-diagnosing, and self-healing IT estates on-premises and in a cloud environment through a modular plug-and-play approach. This modular approach is important for supporting enterprises' existing brownfield AI and automation investments. For example, TCS' Cognix GenAI augmented platform delivers analytics-driven operations while harnessing client investments.

Vendors are deploying IP and third-party tools and platforms to support clients' AIOps and automation strategies. For example, Kyndryl is enhancing its Kyndryl Bridge AIOps and intelligent automation capabilities to drive touchless self-healing systems across incident prevention, auto resolution, policy-based guardrails, and AI/ML driving insights and recommendations. Wipro, for example, has developed an AIOps and observability framework, including full-stack observability, to help clients accelerate their zero-ops journey.

There will be increased utilization of AI copilots to improve engineers' efficiency with AIOps incident prediction and developer experience with GitHub Copilot, including legacy code modernization and the development of functional GenAI use cases, including AI agent assist, AI voice translation, onboarding AI agent, engineer assist, etc. TCS, for example, empowers L1s to resolve issues with data center assist (GenAI knowledge retrieval and insight generation for data center-related user queries).  Investment in AI-enabled copilots will continue providing users with intelligent suggestions, automating routine tasks, and providing real-time insights to improve efficiency.

We expect to see an expansion of small language models to meet industry-specific requirements, such as a GenAI-powered advisor built on local LMs customized to client data. In addition, a single delivery team for GenAI in key centers and CoEs. There will be more focus on Agentic AI frameworks to enable citizen developers to create agentic workflows and agents to collaborate with human agents. GenAI will be used to drive more experience and workflow orchestration, moving further up the stack for full business application-level visibility and improving business outcomes. In addition, it accelerates automation, change, insight generation, and problem-solving capabilities. 

There will also be increased investment in private AI, including NVIDIA stack for on-premises GenAI.  For example, Infosys helps clients set up NVIDIA DGX Cloud, providing a tested and certified blueprint deployed and managed as a hosted or client on-premises model. TCS provides HPE Private Cloud AI with embedded NVIDIA AI capabilities, and DXC Technology is standardizing NVIDIA technology in deploying enabling infrastructure for clients to run AI/GenAI projects.

Increasing demand for SRE-led operations to improve reliability & experience

Vendors are expanding dedicated SRE and DevSecOps practices and resources, and an SRE-led approach to cloud operations and full-stack observability capabilities, with SRE-based operations reducing operations toil through an automation-first approach. This includes expanding SRE advisory services to define an SRE maturity roadmap supported by SRE assets and accelerators; also, an SRE adoption framework, transformation services, process framework, and SRE tools and best practices. Recent vendor examples across SRE include:

• NTT DATA has developed SRE Agile Pods for designing, deploying, managing, and automating technology assets, including autonomous CI/CD pipeline application artifact delivery or building complete infrastructure as code
• TCS is empowering SREs to deliver end-to-end reliability and focusing more on AIOps and remediation
• Kyndryl is rolling out SREs across all its mature accounts and has altered its career model to recognize SRE as a defined role
• Infosys has a structured framework to enable clients to progress towards SRE-based operations and reduce operations toil with an automation-first approach
• Coforge has shifted to an SRE-centric way of working and upskilling teams to become problem solvers with an engineering mindset
• Unisys is ramping skillsets across AI, including fully adopting the SRE model.

Greater focus on OCM & AI strategy & consulting services to drive technology adoption & ROI

AI has a much wider role to play than supporting cognitive IT infrastructure. For example, vendors are adopting a cognitive consulting and an advisory-led approach to expedite clients' AI transformation strategies. This includes taking a collaborative design thinking approach and utilizing IP and frameworks to co-create and co-innovate with clients on their AI journeys. This encompasses discovery workshops and automation and AI maturity analysis; in addition, tools assessment, strategy, and roadmap development from AI prototype to full-scale deployment. For example, DXC Technology provides industry-specific GenAI consulting to accelerate adoption across industries. Likewise, Infosys provides AI consulting, assessment, advisory, and implementation services. It further embeds AI and GenAI use cases in all infrastructure deals in POC and production.

Clients are looking to utilize existing tooling investments and enable the orchestration of tools, including AI, through a single pane to support their business outcomes. Another key focus is AI as a service where clients are looking for help with AIOps and benefits, as well as AI-foundation setup and management of the change from a development and overall consumption standpoint. There is an increasing focus on OCM to support adoption and integration through AI. Unisys, for example, provides AI-enabled OCM and persona-based training to expedite adoption rates and improve overall experience.

Over the next 12 months, vendors will focus on persona-based, AI-enabled OCM to drive the adoption of copilot and AI capabilities, for example, to increase ROI across the enterprise. This includes device and sentiment insights to inform training methodologies and technology adoption rates.

We expect to see a greater focus on the SRE model, increasing engineers' productivity with AI-assisted steps for resolution and shortening the learning curve across infrastructure operations.  Also, there will be increased seeding of SREs into client end-to-end teams and building competencies over time. This includes GenAI use cases, Agentic AI, small LMs, and algorithms for AIOps platforms, among others, aligning SREs to verticals with value streams and domain skills as clients move to product-centric models.

Outlook

Investment in cognitive & self-healing IT infrastructure management services will continue to ramp up with GenAI as a service offering, and GenAI will enable new product lines. This includes building GenAI capabilities with a native stack and Agentic AI use cases supporting AI-ready infrastructure.  

There will be more focus on DEX to drive holistic experience across the enterprise and measure total experience through AI-enabled unified monitoring and observability, in addition to advanced AI integration for predictive analytics and anomaly detection.

Vendors need to expand AI platforms orchestration, AI-enabled monitoring, and SRE frameworks and adoption.