Cisco is rolling out AI agents to every single one of its 90,000 employees

Good morning. Mark Patterson has spent 26 years at Cisco Systems, which is long enough to watch the company navigate multiple technology cycles. But he says nothing compares to what’s happening now. “AI is the most significant technology transition that we’ve seen in probably our lifetime, and I think Cisco was right at the heart of it,” Patterson, CFO since July 2025, told me. That includes not just building the infrastructure powering the AI economy but embedding AI into how the business runs. Starting in its new fiscal year at the end of July, Cisco (No. 83 on the Fortune 500) is rolling out AI agents to its approximately 90,000 employees. Each employee will have access to a personalized assistant capable of handling tasks, answering questions, and routing requests to the most efficient AI model. The company said it does not separately disclose those costs outside of its earnings reporting. The architecture reflects a focus on performance optimization. “It’s not going to burn a whole bunch of tokens with frontier models,” Patterson said. Cisco’s system dynamically selects the right model for each task. “It knows which tool is most effective and most efficient,” he said. Much of the infrastructure is built on-premises, giving Cisco greater control over cost and data. “We feel like that’s the most efficient way is to build our own AI stacks, which will go out and query the different models based on the particular use case,” he said. In general, AI agents use far more tokens than standard chats because they continuously plan, call tools, and process intermediate steps. While a typical chat may use a few thousand tokens, complex agent tasks can consume hundreds of thousands to millions in a single run. Patterson has his own agent as well. He uses it primarily for benchmarking—quickly comparing Cisco’s performance against peers across metrics like revenue growth, EPS, R&D spend, and capital allocation, often throug