Dun & Bradstreet Rebuilds 642 Million Business Database for AI Agents

Quick Facts

• D&B rebuilt its Commercial Graph covering 642 million businesses after discovering AI agents couldn’t work with systems designed for humans
• The company now processes 100 billion data quality checks monthly and migrated to cloud infrastructure with sub-second query response times
• D&B launched three AI agents for KYC workflows, credit analysis, and sales preparation, integrated with Anthropic’s Claude via Model Context Protocol

Dun & Bradstreet spent months rebuilding its massive business database after discovering that AI agents couldn’t work with systems designed for human analysts.

The 164-year-old data company’s Commercial Graph tracks 642 million businesses across 200 countries. The database nearly doubled in five years, expanding from 300 million records to its current size, with 11,000 fields per business record.

The problem emerged when D&B’s customers began deploying AI agents for credit assessments, procurement, and supply chain workflows. The fragmented database architecture couldn’t deliver the sub-second response times agents require.

“Querying that at the sub-second latency agents require, against a fragmented architecture, was not workable,” said Gary Kotovets, D&B’s chief data and analytics officer. “The systems built to serve human analysts were the wrong architecture for machines.”

D&B migrated its fragmented databases to cloud infrastructure and built a data fabric layer that normalizes records across markets. The company now processes 100 billion data quality checks monthly as records move through its systems.

The rebuild created a unified knowledge graph that tracks billions of relationships across all 642 million companies. D&B built a structured access layer on top that packages data with context and routes agents to verified, specific entities rather than simple name matches.

“We need to think about agents as our new consumer category,” Kotovets told VentureBeat. The D-U-N-S number serves as “the connective tissue between the multiple agents,” allowing systems to reference the same entity as information moves across platforms.

D&B launched its D&B.AI suite with three deployed agents for KYC workflows, credit analysis, and sales call preparation. The platform includes ChatD&B for natural language queries and Model Context Protocol servers for standardized data access.

In May 2026, D&B announced integration with Anthropic’s Claude, bringing D&B risk data directly into the AI assistant to speed business onboarding and compliance work.

“The next wave of AI is agentic, and it runs on trusted data,” said CEO Steve Tulenko. “Our capabilities become a platform your people and your agents can build on with confidence.”

The transformation addresses a fundamental shift in enterprise AI adoption. Multi-agent systems grew 327% in four months, with more than 80% of databases now built by AI agents.

D&B serves nearly 200,000 customers globally, including 90% of Fortune 500 companies. The company’s data is already embedded in critical business workflows, making the transition to agentic AI applications smoother for existing customers.

Read more: D&B’s database of 642 million businesses was built for humans, not AI agents. So they rebuilt it.