The urgent need to ship and monetize agentic AI features has turned the Chief Product Officer job from a strategic planning role into a high-pressure, execution-focused position, arguably the most stressful in B2B tech today.
Rubrik's CPO asserts that vendors are blamed for—and are ultimately responsible for—customer errors. The vendor's duty is to build agents with guardrails that make correct choices obvious and prevent mistakes, shifting the accountability model for enterprise software.
Glean's strategy extends beyond being a user-facing app. It positions itself as a centralized system of intelligence that provides superior, offline-processed context to other AI front-ends like Claude or Cursor, a rapidly growing use case.
Legal AI company Harvey uses specialized deployment pods that include "legal engineers"—who are former lawyers—alongside software engineers and PMs. This model embeds deep domain expertise directly into the customer onboarding and customization process for high-stakes verticals.
Cybersecurity company Rubrik aims to transition its entire product to an agentic interface. The goal is for customers to operate all platform functions through conversational commands, making traditional UIs and dashboards obsolete in complex enterprise software.
Contrary to stereotypes, Harvey's CPO finds lawyers at top firms are not tech laggards but are "incredibly curious" and driven by a "competitive spirit" to adopt agentic AI. Their focus is on collaborating with and verifying the AI's output.
Glean discovered an unexpected growth loop: as usage of its intelligence layer *inside* other platforms like Claude grows, it creates a "spillover effect" that brings users back into the native Glean UI to build more powerful, team-wide agents and assets.
For high-stakes tasks, fully autonomous agents are too risky. The effective model, used by cybersecurity firm Rubrik, is for the AI to generate a detailed plan of action, which a human expert then reviews, edits, and approves before execution.