How are you helping companies break down silos with these, these new teammates?

Yeah.

So my personal philosophy, and this is my personal philosophy, my personal philosophy with AI agents is that they should act on behalf of a user, not as a user.

And the distinction there is, I believe from a security segregation of duties, uh, perspective, we want AI agents to act as Michael, not as itself as, like, some sort of super agent with super access that can do anything.

Uh, AI agents should act on behalf of the user that they're supporting and adopt that user's authorities so that, you know, over the last decade, 20 years, we've put a lot of-- we organizations have put a lot of effort into defining security roles or segregation that have aligned with SODs, segregation of duty.

So for example, I can't create a vendor, receive an invoice, pay an invoice, right? No one person can do that 'cause theoretically you could set up your own vendor and pay yourself, and then no one would know, right? So AI agents shouldn't be able to do that either.

They should be able to act on behalf of users, and we help customers define not only the tasks that an AI agent should be supporting, but the security architecture, uh, the monitoring, the governance of AI, in addition to the implementation of AI because the operas- the to operationalize AI agents beyond the conference room pilot is actually pretty tough.

Would love to hear about, you know, some of the, the toughest lessons or most interesting lessons you've learned, you know, as you transitioned across those different roles a-and now working within a, a global organization.