Vercel has made itself a test case: The company is productively deploying AI agents across nearly every business function – from inbound sales to customer support. One AI agent already answers 91 percent of all support tickets, while another handles lead qualification. Julian Eckerle, Head of Central Europe at Vercel, shared these insights in the podcast "t3n Arbeit in Progress."
The Key Facts
- 91 percent of support tickets are handled by an AI agent
- AI integration began approximately two and a half years ago with a bottom-up approach
- Only 11 percent of German companies regularly deploy AI agents; 60 percent are planning or discussing it
- No layoffs so far – jobs have been reshaped, not eliminated
From Chaos to Governance
Vercel's strategy was initially radical: the company asked all employees to build their own AI agents. The goal was to foster innovation at the organization's edges. Eckerle explains the thinking: "We wanted to experiment with what happens when we make these tools available to the outer reaches of the organization to drive innovation."
This quickly generated a large market of use cases – agents for content, support tickets, lead processing. But control threatened to slip away. So Vercel introduced a governance structure to track token consumption, enforce security policies, and align everything with the identity provider.
"About eleven percent of German companies regularly deploy AI agents today. Sixty percent are planning or discussing it. And when you see how many agents actually deliver real value in the end, we're still in a relatively early phase." — Julian Eckerle, Head of Central Europe at Vercel
Automation, Not Elimination
Here's what's striking: Vercel claims to have conducted no layoffs so far. Instead, roles were reshaped. Support staff now handle complex cases the agent cannot resolve. Sales teams focus on strategic accounts rather than manual lead qualification.
This differs from the typical "AI replaces humans" narrative – at least according to Vercel's account. Whether the model remains sustainable or roles eventually disappear remains an open question.
What This Means for German Companies
Vercel's example shows that AI agents are no longer science fiction but operationally deployed – if the infrastructure is right. For German mid-market and enterprise companies, this could mean: those who don't start systematically testing AI agents now risk falling behind. At the same time, Vercel's experience warns against uncontrolled growth without governance – token costs, security, and compliance must be considered from day one. The question is less "AI yes or no" and more "How do we organize this at scale?"
Sources
Editorially owned by Ideal Syka. Sources and method: Newsroom & method. Tips and corrections: ai@i6eal.de.
