What does it actually take to move beyond AI pilots and turn enterprise ambition into real productivity gains?

That question sat at the center of my conversation with Olivia Nottebohm, Chief Operating Officer at Box, and it is one that every boardroom seems to be wrestling with right now. AI conversations have matured quickly.

The early excitement has given way to harder questions about return, trust, and what changes when software stops assisting work and starts acting inside it.

Olivia brings a rare vantage point to that discussion, shaped by leadership roles at Google, Dropbox, Notion, and now Box, where she oversees global go to market, customer success, and partnerships at a time when AI is becoming embedded in everyday operations.

We talked about why early adopters are already seeing productivity lifts of around thirty seven percent, while others remain stuck in experimentation. The difference, as Olivia explains, is rarely the model itself. Strategy matters more.

Teams that treat AI as a chance to rethink how work flows through the organization are pulling away from those that simply layer automation on top of broken processes. This is where unstructured content, often described as dark data, becomes a competitive asset rather than a liability. When that information is curated, permissioned, and ready for agents to use, entire workflows start to look very different.

A large part of our discussion focused on AI agents and why 2026 is shaping up to be the year they move from novelty to necessity. Agents are already joining the workforce, taking on tasks that used to require multiple handoffs between teams. That shift brings speed and autonomy, but it also raises new questions about trust.

Olivia shared why governance has become one of the biggest blind spots in enterprise AI, especially when agents act independently or interact across platforms. Her perspective was clear. Without strong security, permissioning, and oversight, the risks grow faster than the rewards.

We also explored why companies using a mix of models and agents tend to see stronger returns, and how Box approaches this with a neutral, customer choice driven philosophy while maintaining consistent governance.

From the five stages of enterprise AI maturity to the idea of a future agent manager role, this conversation offers a grounded look at what AI at scale actually demands from leadership, culture, and operating models.

So as investment accelerates and AI becomes part of the fabric of work, the real question is this. Are organizations ready to redesign how they operate around agents, data, and trust, or will they keep experimenting while others pull ahead, and what do you think separates the two?

Useful Links

Connect with Olivia Nottebohm
The State of AI in the Enterprise Report
Becoming an AI-First Company
Follow on LinkedIn

Thanks to our sponsors, Alcor, for supporting the show.

3546: Box and the Leadership Shifts Behind Becoming an AI First Company

Tech Talks Daily