Cisco NetworkingCisco Nexus DashboardCisco AI SolutionsCisco LiveCisco Data Center NetworkingCisco
Blog • • 4 minute read

Cisco AI Canvas: Transforming IT Operations with Generative AI and AgenticOps

Cisco AI Canvas is a generative AI workspace for IT operations, built on Cisco's Deep Network Model. Designed to support AgenticOps, it enables faster issue resolution, automation and collaboration across networking, security, cloud and observability. Learn how Cisco is transforming infrastructure management with AI.

In this blog

  1. What Is AgenticOps?
  2. What Is Cisco AI Canvas?
  3. Key Features That Differentiate Cisco AI Canvas
  4. Cisco AI Assistant: Contextual, Conversational and Integrated
  5. Deep Network Model: Cisco's Purpose-Built AI Engine for Networking
  6. A New Paradigm in IT Operations
  7. Early Access and Roadmap

As artificial intelligence reshapes the technology landscape, Cisco introduces a breakthrough innovation in IT operations: Cisco AI Canvas. This powerful solution is a core component of Cisco's AgenticOps strategy, designed to modernize how IT teams manage, troubleshoot, and automate across the network stack. With Cisco AI Canvas, generative AI meets deep networking expertise to deliver an intelligent, collaborative workspace for the future of IT.

 What Is AgenticOps?

AgenticOps is Cisco's next-generation operational model that combines human expertise with AI-driven automation to improve how digital infrastructure is managed. It is built around three core principles:

  • Human-in-the-loop control for trusted oversight
  • Cross-domain context to eliminate data silos and unify operations
  • Purpose-built agents trained to take action in complex IT environments

By combining these elements, AgenticOps empowers intelligent agents to reason, act, and learn continuously, accelerating problem resolution, improving operational efficiency, and scaling IT productivity.

 What Is Cisco AI Canvas?

Cisco AI Canvas is a shared, generative AI workspace designed to help IT teams collaboratively resolve issues, automate tasks, and make real-time infrastructure decisions. Built specifically for modern enterprise environments, AI Canvas spans networking, security, cloud, observability, and beyond.

By integrating with the Cisco AI Assistant and leveraging the Deep Network Model, Cisco AI Canvas provides dynamic, data-rich insights tailored to each issue. This makes it faster and easier for teams to identify root causes, propose solutions, and apply remediations in a unified view.

 Key Features That Differentiate Cisco AI Canvas

1. Built on Cisco's Deep Network Model
Cisco AI Canvas is powered by the Deep Network Model, a specialized large language model trained on over 40 years of Cisco expertise. Unlike general-purpose AI, this model is purpose-built for networking and infrastructure, enabling highly accurate diagnostics, configuration analysis, and remediation.

2. Agentic Intelligence for IT Operations
AI Canvas introduces agentic intelligence—AI agents that don't just assist, but act autonomously based on context. These agents can recommend or execute changes, generate dynamic runbooks, and continuously learn from outcomes to improve over time.

3. Unified Cross-Domain Visibility
Troubleshooting across firewalls, apps, and networks used to require multiple dashboards. Cisco AI Canvas consolidates observability data from tools like ThousandEyes, Meraki, and Splunk into a single, real-time view, making it easier to correlate issues and streamline workflows.

4. Seamless Collaboration Across IT Roles
From Tier 1 support engineers to NetOps, SecOps, and IT leadership, AI Canvas enables real-time collaboration, shared dashboards, and persistent workspaces. Teams can escalate, delegate, and resolve issues together—all within a single platform.

5. From Diagnosis to Execution
Unlike traditional observability tools, Cisco AI Canvas guides users from initial alert to solution execution. Teams can implement AI-suggested fixes directly within the platform or roll them back as needed, with human oversight embedded throughout.

 Cisco AI Assistant: Contextual, Conversational and Integrated

At the heart of Cisco AI Canvas is the Cisco AI Assistant—a conversational interface that uses natural language to interpret questions, retrieve relevant data, and generate action plans. Whether it's "diagnose packet loss in St. Louis" or "check firewall policy impact," the AI Assistant works alongside the Deep Network Model to deliver precise and fast insights.

 Deep Network Model: Cisco's Purpose-Built AI Engine for Networking

Unlike generic LLMs, the Cisco Deep Network Model is engineered for infrastructure-level tasks. It has been trained on:

  • Over 40 million tokens of Cisco networking knowledge
  • 3,000+ annotated reasoning paths
  • Cisco TAC and Customer Experience case data
  • Real-world configurations, topologies, and logs

The model also leverages reinforcement learning, enabling continuous improvement based on real-world usage. With a compact architecture (ranging from 8 to 30 billion parameters), it can even operate in edge and air-gapped environments, ensuring flexibility and scalability.

 A New Paradigm in IT Operations

Cisco AI Canvas brings the AgenticOps vision to life. In this new paradigm:

  • Human oversight is embedded in every interaction
  • AI agents collaborate across domains, breaking down silos
  • Specialized, purpose-built models drive deeper accuracy and performance

Together, AI Canvas and AgenticOps redefine how IT organizations resolve problems, manage infrastructure, and deliver business outcomes.

 Early Access and Roadmap

Cisco AI Canvas will be available in early access starting in August 2025, allowing customers to co-develop and provide feedback on capabilities. This agile approach ensures the platform evolves with real-world needs and continues to drive value across the enterprise.  Contact your WWT Account Manager to get started with Cisco AI Canvas.

Technologies