Particle.news

Cisco Positions Splunk as Enterprise Control Plane for Agentic AI

Shipping an agent-first control plane with observability, Tokenomics cost tools, expanded on-prem NVIDIA POD support, and a reported multi-year AWS pact, Cisco aims to give firms clear visibility, budget control and stronger security for continuously running AI agents.

Overview

  • At its customer conference in Denver, Cisco and Splunk announced product releases that turn prior roadmaps into shipping offerings designed to manage agentic AI through a central control plane.
  • Cisco Cloud Control is presented as an agent-first centralized control plane that unites telemetry and security to track agent activity, model performance and runtime across the stack.
  • New Tokenomics tools will track token consumption in real time, assign costs to teams and forecast spend so organizations can link AI usage to budgets and business metrics.
  • Cisco expanded on-prem options with pre-validated Cisco AI PODs using NVIDIA compute for self-managed, air-gapped deployments and said it plans support for additional local models.
  • The companies also outlined an agentic SOC, broader observability features and a reported multi-year AWS product development agreement to speed cloud-scale security integration, building on Cisco’s 2024 acquisition of Splunk.