From Hype to Baseline: Agentic AI Enters Core Enterprise Workflows
Agentic AI is moving from slideware to production in some of the most conservative corners of the enterprise stack. Platforms that run daily operations—IT service management, project work, HR and finance workflows—are embedding autonomous agents directly into their core products. Workday is extending its Sana platform from HR and finance into IT service management, while ServiceNow has used its own agentic AI internally to streamline complex processes. Atlassian has made agents first-class citizens inside Jira, and GoTo is turning AI into a strategic partner in its LogMeIn Resolve and Rescue tools. The result is a clear architectural shift. Instead of chatbots bolted onto ticketing queues, organizations are experimenting with AI service management built into systems of record, where agents can see policies, roles, historical tickets, and context. For IT leaders, the question is no longer whether agentic AI will touch enterprise workflows, but which platforms will orchestrate it—and how tightly it should be integrated into IT operations.
ServiceNow’s Self-Experiment: Eight Seconds Instead of Four Days
ServiceNow offers one of the clearest measurable proofs of what happens when agentic AI is applied to real enterprise complexity. Rather than limiting new capabilities to customer demos, the company ran agentic AI on its own internal processes. One standout use case is a redesigned commissioning workflow for sales employees. Previously, queries had to be submitted to finance, with an average turnaround time of four days. The AI-enabled version, built with guardrails around security and policy, now resolves the same query in eight seconds. This is not just a marginal improvement; it changes the operating model. Decisions that previously stalled deals now complete in near real time, and manual handoffs are replaced by AI-driven orchestration. For CX and IT leaders evaluating IT operations AI, the key takeaway is that agentic AI can deliver tangible enterprise automation results—but only when integrated into the core process and governed with clear controls, not treated as a superficial chatbot layer.
Workday and GoTo: Embedding AI Service Management Into Daily IT Work
Workday’s launch of Sana for IT Service Management shows how agentic AI can reduce the need for traditional tickets altogether. Because it runs on Workday’s existing policies and security model, the agent already knows an employee’s role, manager, and approval chain. Routine tasks like password resets, software installs, and access changes can be resolved conversationally, without touching a separate ITSM tool. More importantly for IT operations, lifecycle events such as hiring, role changes, and offboarding become triggers for automatic provisioning and deactivation across identity, security, and collaboration systems. GoTo’s enhancements to LogMeIn Resolve point to a similar direction, focused on the front line of support. Its agentic AI powers a Resolution Agent that interprets user requests, asks clarifying questions, runs diagnostics, and proposes fixes that technicians can approve with one click. Customers report faster knowledge base creation and fewer manual steps, translating into concrete efficiency gains in day-to-day support and troubleshooting.
Jira’s AI Assignees: Agents as First-Class Team Members
Atlassian has taken a different but complementary route, weaving agentic AI directly into task and project management. In Jira, AI agents are now assignable resources alongside human team members, subject to the same permissions, audit trails, and governance. Tasks can be assigned directly to Rovo, Atlassian’s native agent, or to third-party agents connected through the Model Context Protocol, which links Jira to tools like Figma, GitHub, Canva, Box, and Intercom. Usage data suggests that this is not a niche experiment. Atlassian reports that agentic automation runs across its platform are growing 30% month over month, with Model Context Protocol usage doubling at a similar rate. Rovo Dev handles repetitive developer tasks such as security patches and dependency changes with human approval, while Rovo Service executes employee support workflows by pulling from past tickets, knowledge bases, and policies. This shows agentic AI shifting from a helper to an accountable owner within IT operations AI and project delivery.
What the Early Numbers Really Say About ROI and Risk
Across these deployments, a pattern is emerging in agentic AI enterprise adoption. When implemented as part of the system of record—rather than as an isolated bot—agents deliver measurable efficiency improvements. ServiceNow cites a cut from four days to eight seconds in a core sales-finance process. Atlassian’s data shows rapidly compounding usage, and customers using its Rovo agent are seeing faster recurring revenue growth. Workday and GoTo both illustrate how integrated AI service management can shrink ticket volumes by resolving issues at the source and automating lifecycle-driven actions. However, the same projects underline why success is not guaranteed. Each vendor stresses guardrails: one-click approvals before LogMeIn Resolve executes fixes; human sign-off before Rovo ships code; embedded policy and security models in Sana’s ITSM agent. The practical lesson for IT leaders is clear. Real enterprise automation results depend less on how powerful the model is and more on how deeply AI is wired into workflows, data context, and governance.