# Univé Proves Enterprise AI Needs Builders
_Univé is scaling AI by turning 1,500 employees into custom tool builders, cutting insurance claim times from hours to minutes._
**Published:** 2026-07-31
**Source:** https://www.startuphub.ai/ai-news/artificial-intelligence/2026/univ-proves-enterprise-ai-needs-builders
---
Dutch insurer Univé has shifted its corporate strategy to prioritize internal AI development, moving beyond simple tool deployment to build a workforce of active creators. According to [OpenAI News](https://openai.com/index/unive), the company now relies on 1,500 custom GPTs to handle daily operations across finance and legal departments.

Univé's AI StrategyDriver

From the article 2 mentionsDutch insurer Univé has shifted its corporate strategy to prioritize internal AI development, moving beyond simple tool deployment to build a workforce of active creators.leads to1,500 Employee BuildersCoreturning 1,500 employees into custom tool builders, not just users of AICustom GPTsCoreFrom the articleAccording to OpenAI News, the company now relies on 1,500 custom GPTs to handle daily operations across finance and legal departments.Competitive EdgeOutcomeFrom the articleUnivé director Yous van Halder suggests the competitive edge lies in the volume of employees learning to build rather than the tools themselves.enablesAgentic WorkflowsContexttransition toward proactive task preparation, gathering data before human reviewFrom the article 2 mentionsThe transition toward agentic workflows in enterprise AI is defined by proactive task preparation.Faster Claim TimesOutcomecutting insurance claim times from hours to minutes with automated data prepFrom the articleBy the time a claims professional begins their review, the heavy lifting is complete.Permission InheritanceContextsoftware only accesses data the individual user is already cleared to seeFrom the articleThis shift to OpenAI Workspace Agents implementation relies on strict permission inheritance.

## The Shift to Agentic Workflows

The transition toward agentic workflows in enterprise AI is defined by proactive task preparation. Rather than waiting for a request, these systems gather invoice data and policy documents before a human handler even opens a file. By the time a claims professional begins their review, the heavy lifting is complete.

This shift to [OpenAI Workspace Agents implementation](/ai-news/artificial-intelligence/2026/chatgpt-workspace-agents-admin-builder-controls) relies on strict permission inheritance. The software only accesses data the individual user is already cleared to see, preventing security slips.

Univé director Yous van Halder suggests the competitive edge lies in the volume of employees learning to build rather than the tools themselves. This philosophy marks a departure from traditional IT projects that rely on centralized engineering teams to solve every minor bottleneck.

For founders, this signals a shift in enterprise procurement. Large organizations are moving away from buying rigid black-box software in favor of platforms that allow their own staff to customize workflows.

Underwriting queues now arrive pre-structured with risk indicators and missing documentation flagged in advance. The human expert retains the final decision, but the time spent on manual data assembly has vanished.
---
Original analysis from [startuphub.ai](https://www.startuphub.ai), the #1 AI startup directory.