The Future of IT Operations: AgenticOps
From AIOps to AgenticOps: a necessary evolution
New realities require new responses. This principle applies even more strongly in information technologies, where constant transformation is the norm.
An assessment of today’s telecommunications network management reveals a scenario of increasing complexity. In this context, organizations need responses that are more autonomous, predictive, and scalable.
Traditional automation is no longer sufficient. AI-powered IT operations platforms (AIOps) represented a major step forward by introducing advanced analytics, event correlation, and machine-learning-based recommendations.
However, most AIOps approaches rely on rule-based automation or closed ML models. As a result, their ability to reason explicitly about context, plan complex actions, or adapt dynamically to new situations remains limited.
The next evolutionary step is AgenticOps. In this model, systems do more than detect and recommend. They reason about problems, decompose goals, and plan and execute actions in a contextualized way. All of this happens at machine speed and within enterprise-grade governance frameworks.
AI agents: from assistants to collaborators
From this new perspective, AI agents are autonomous software entities specialized by role and context. They do not only respond. They decide and act.
Unlike chatbots or dashboards that merely display information, agent-based architectures understand the goals and context defined by humans. They plan and execute tasks, adapt to changing conditions, and continuously improve their behavior.
This improvement is driven by feedback, working memory, and policy refinement. Over time, agents become more effective and better aligned with operational objectives.
The five defining attributes of AgenticOps
Agent-based architectures for IT operations are defined by five core attributes:
1. Identity and context
Each agent operates with a clearly defined role, purpose, and scope.
2. Reasoning
Agents break down complex problems, evaluate alternatives, and make context-aware decisions.
3. Scalability
They operate continuously in always-on, enterprise-scale environments.
4. Security
All actions are governed by policies, permissions, and audit logs.
5. Operational efficiency
By combining reasoning with automation, agents reduce manual effort and accelerate outcomes.
These attributes turn agents into true collaborators. They work alongside other agents and IT teams to power modern operations.
Multidomain IT operations in real time
The result is a new type of multidomain operation. Problems are prevented or resolved faster. Ticket volumes decrease. IT teams can adapt in real time as conditions change.
Operations become more resilient, proactive, and aligned with business needs.
Cisco’s perspective
Within the Cisco ecosystem, this approach is not a future vision. It is already becoming reality across multiple domains through the combination of AI Assistant and AI Canvas.
AI Assistant provides a natural-language interface that helps diagnose issues and automate tasks in minutes. It simplifies collaboration across IT teams and reduces time spent searching for information. This allows teams to focus more on decision-making.
AI Canvas complements this capability. It offers a collaborative surface with intelligent, dynamic dashboards that enable teams to detect, visualize, and resolve problems in real time and in context.
Together, these tools create a shared space where people and AI agents collaborate. AI becomes embedded directly into operational workflows and becomes an active part of the process.
Cisco’s approach integrates telemetry, intelligence, and collaboration into a single, coherent framework. The goal is to make agents enterprise-grade. They must reason with context, act securely, and collaborate seamlessly with human operators. Trust and transparency are built into problem resolution from the start.
An architecture designed for AI
None of this would be possible without an evolved network architecture designed specifically for AI-driven operations.
In June 2025, Cisco introduced a secure network architecture to accelerate AI transformation in the workplace. It was designed to support future campuses, branch offices, and industrial networks.
This architecture delivers unprecedented operational simplicity. It combines unified management, next-generation network devices engineered for predictable latency, and advanced security capabilities embedded directly into the network.
The network is not just a connectivity or performance enabler. It is a foundational component of the AgenticOps model.
It provides AI agents with rich telemetry, real-time context, and predictable latency. These elements are essential for precise reasoning and action. By embedding observability, segmentation, security, and control into the network—including infrastructures optimized for AI workloads such as RoCEv2—automated decisions can be executed securely and traceably.
As a result, the network evolves from a reactively managed domain into an active, context-aware system. It becomes ready to support IT operations at the speed today’s environments demand.
Cisco therefore sets a new standard. Organizations can address traffic growth, rising cyberthreats, and the need for continuous availability while fully harnessing AI’s potential in the workplace.