Cisco Cracks Code on Controlling AI Expenses with Unified Operating Model

The increasing adoption of artificial intelligence (AI) in enterprises has brought about a new challenge - managing AI spend. According to Cisco, IT leaders are struggling to justify AI expenses to finance teams due to a lack of visibility and trust in AI operations.

Cisco's Vice President of Engineering, Richard Delisser, and Vice President of Enterprise AI Platforms and Infrastructure, Greg Sylvester, acknowledged that the company initially faced the same issue. However, they decided to tackle it head-on by building the necessary visibility to control AI spend.

The key to this is 'tokenomics,' which measures efficiency in AI operations. Tokenomics aims to maximize 'token yield' - the return on investment per token spent. Cisco's goal is not to reduce AI expenses but to find an equilibrium between cost and value.

To achieve this, Cisco has integrated its infrastructure, security, and observability capabilities to create a unified operating model. This allows for end-to-end visibility into AI activity, enabling the company to attribute AI spend directly to business value.