D&B's database of 642 million businesses was built for human users, not AI agents. So they rebuilt it for AI.

D&B REBUILDS ITS DATABASE FOR AI AGENTS

Dun & Bradstreet (D&B) has long been a cornerstone in the world of commercial data, having spent over 180 years developing a comprehensive database that encompasses 642 million businesses. Historically, this database, known as the Commercial Graph, was meticulously designed for human users such as credit analysts, risk managers, and sales professionals. These users were accustomed to waiting for query results and navigating through ambiguous entity matches, tasks that AI agents struggle with. However, as D&B's clients began integrating AI agents into their credit, procurement, and supply chain workflows, it became clear that the existing architecture was inadequate for machine-driven operations. In response to this pressing need, D&B undertook a significant transformation, rebuilding its database to cater specifically to AI agents.

THE CHALLENGES D&B FACED WITH HUMAN-CENTRIC DATABASE DESIGN

The original design of D&B's database presented several challenges when it came to accommodating AI agents. The Commercial Graph was not a singular database but rather a collection of disparate systems tailored for various use cases and markets. This fragmentation was manageable for human analysts, who could utilize SQL queries or pre-built interfaces to navigate the complexities of the data. However, AI agents, which require streamlined and efficient access to data, found this structure to be a significant barrier. The inability of these agents to handle the ambiguity that human analysts could navigate meant that the existing database design was fundamentally misaligned with the needs of modern AI technologies.

HOW D&B'S COMMERCIAL GRAPH ADAPTED TO AI QUERYING NEEDS

Recognizing the limitations of their previous architecture, D&B initiated a comprehensive overhaul of the Commercial Graph to better serve AI querying requirements. This involved rethinking the underlying data structures and integration methods to ensure that AI agents could efficiently access and interpret the vast amounts of information contained within the database. By focusing on creating a more cohesive and accessible data environment, D&B aimed to eliminate the fragmentation that had previously hindered AI performance. This transformation not only improved the querying capabilities for AI agents but also enhanced the overall usability of the database for all users, blending human and machine needs into a unified framework.

THE IMPACT OF AGENTS ON D&B'S DATABASE ARCHITECTURE

The shift towards accommodating AI agents necessitated a fundamental change in D&B's database architecture. As the company moved from a human-centric design to one that prioritized machine interaction, it became clear that the scale of the data presented both challenges and opportunities. The database had nearly doubled in size over five years, growing from more than 300 million to over 642 million business records. This exponential growth meant that the new architecture had to be robust enough to handle not only the volume of data but also the speed and efficiency required by AI agents. The integration of advanced querying capabilities and streamlined data access points was essential to ensure that agents could operate effectively within the newly designed system.

LESSONS LEARNED FROM D&B'S TRANSFORMATION FOR AI USAGE

D&B's transformation offers valuable insights into the complexities of adapting legacy systems for AI usage. One key lesson is the importance of designing databases with both human and machine users in mind. The previous architecture, while effective for human analysts, proved inadequate for the demands of AI agents, highlighting the need for a more integrated approach. Additionally, the rapid growth of data necessitates ongoing evaluations of database structures to ensure they can support evolving technologies. D&B's experience underscores the necessity of flexibility and adaptability in data management, particularly as AI continues to play a more prominent role in commercial workflows. As organizations look to the future, the lessons learned from D&B's journey will be crucial in shaping the next generation of database design and functionality.