Archive — US — techcrunch
Vercel CEO Guillermo Rauch on the fight to split off models from agents
Guillermo Rauch, CEO of Vercel, discusses the evolution and challenges of AI agents within the company after their recent ShipNYC conference. Vercel, with 6 million daily deployments, emphasizes the need for secure AI data handling and developing frameworks like Eve and Vercel Sandbox to enhance productivity and data control while competing with major AI labs for technological advancement.
Key points
- Vercel is experiencing 6 million deployments daily, with half driven by coding agents.
- The company developed key frameworks, Eve for agent instructions and skills, and Vercel Sandbox for data control.
- Rauch emphasizes the importance of securely managing AI-generated data while allowing agents to operate effectively.
- Internal agents facilitate tasks traditionally bottlenecked by data retrieval, increasing organizational efficiency.
- Client relationships with AI labs are evolving towards plug-and-play models rather than exclusive partnerships.
- Competition arises as AI models evolve, potentially encroaching on Vercel's infrastructure services.
Why it matters
- The rise of AI necessitates robust frameworks for data management, aligning with the increased focus on AI governance in 2025.
- Vercel's push for open protocols mirrors industry trends towards interoperability and modular systems.
- The competition with major AI labs highlights the ongoing trend of democratizing AI tools in tech advancements.
Vercel's ongoing strategy to enhance AI agent functionalities and secure data management positions it at the forefront of evolving AI infrastructure, embodying a significant shift towards open and efficient technology solutions in 2025.