Screaming in the Cloud AI Agents, Enterprise Risk, and the Future of Recovery: Rubrik’s Vision with Dev Rishi
10 snips
Dec 4, 2025 Dev Rishi, GM of AI at Rubrik and a former machine learning CEO, shares insights on enterprise AI adoption and the evolution of agentic systems. He discusses the challenges enterprises face with AI, emphasizing the gap between consumer excitement and organizational risk aversion. Dev introduces Rubrik's innovative Agent Rewind, a safety net for AI-driven actions, helping prevent costly data loss. The conversation also covers trends in AI support, the importance of observability, and the role of governance in ensuring resilience in this rapidly changing landscape.
AI Snips
Chapters
Transcript
Episode notes
Agent Rewind Explained
- Dev Rishi described Agent Rewind as a safety blanket that ties agent actions to backups for one-click recovery.
- He said it correlates agent operations with production changes and restores from a healthy snapshot.
Risk Posture Limits Enterprise AI
- Deep learning and pre-trained models unlocked value for unstructured data like text, images, and video.
- Dev Rishi argued many enterprises still lack the risk frameworks to let AI perform high-value work.
Agents Are Tools With Actions
- Dev defines agents as models with access to tools that can take actions in systems.
- He warned non-deterministic agents granted write access create significant enterprise risk and hesitation.