Chapter 5. Modernize Your Information Architecture

Many companies have a problem with legacy software: mission-critical software that was written decades ago. However, legacy software is only the tip of the iceberg; the more dangerous part lies below, and that’s legacy data. An AI transformation can’t be completed unless you modernize your information architecture. That modernization is the foundation on which the AI Ladder rests (Figure 5-1). As you will come to realize (and already know from your personal lives), when climbing a ladder, foundation matters. A solid foundation lets you climb higher, faster, and with more confidence.

Figure 5-1. A modern information architecture is the foundation on which the AI Ladder sits

What’s the problem that needs to be solved? Why is legacy data an issue? Too often, data is stored in departmental silos: every department has its own database (and these databases often contain the same data at varying degrees of staleness), its own rules for who can access it, and its own definitions for what’s in the database. Working with siloed legacy data is obviously a problem.

Fragmentation isn’t limited to departmental database silos either. Many companies have rightly taken a leap into the cloud, but without addressing the silo issue. That means that their data lives in many clouds: some in IBM’s, some in Amazon’s, some in Microsoft’s or Google’s, ...

Get The AI Ladder now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.