Back to news
AI Business
3d ago

Vercel CEO Discusses AI Development and Competition with Major Labs

Jul 6, 2026
AI Summary

Vercel CEO Guillermo Rauch highlights the company's significant role in AI software, noting 6 million daily deployments and over 1 trillion tokens processed. He discusses the evolution of AI applications, particularly coding and internal agents, and the challenges of data security and integration with existing systems.

  • Vercel is a cloud infrastructure provider that enables developers to deploy agents without server management, currently facilitating 6 million deployments daily, half of which are driven by coding agents.
  • The company processes more than 1 trillion tokens through its AI gateway each day.
  • CEO Guillermo Rauch emphasizes a shift from prototyping to practical applications in AI, focusing on the challenges faced when deploying agents in production environments.
  • Two primary use cases for agents identified by Rauch are coding agents, which drive significant token utilization, and internal agents that assist in company operations, raising concerns about data access and security.
  • Vercel has developed a framework called Eve for defining agent instructions in natural language and a tool named Vercel Sandbox to control data access and ensure security.
  • Rauch notes the risks associated with AI tools potentially accessing sensitive codebases, highlighting the importance of data control.
  • Internal corporate agents, such as those used by sales representatives, can now provide real-time data insights, improving productivity and decision-making processes.
  • The landscape of client relationships with AI labs is changing, with companies now opting for a mix of models and platforms rather than committing to a single provider.
  • Vercel faces competition from major AI labs like OpenAI, particularly as these labs expand their capabilities, which may overlap with Vercel's infrastructure services.
  • Rauch envisions Vercel positioning itself as a leading platform for open protocols in the evolving AI landscape.
vercelguillermo rauchprice performanceproduction optimizationai models