"Michael, I have a simple question." Silvia Seven, CFO of FastChangeCo, pushes the project status deck aside. "How many projects have redefined what a 'customer' is from scratch in the last two years?" Michael Mueller swallows. He knows the answer. And it's uncomfortable.
"Do we actually still need data modelers if AI can take over?" The question came up after a coaching session. I paused. Not because the question was fundamentally wrong — but because it was the wrong question.
Amal leans back and looks at her notebook. Three pages filled. Plus four whiteboards covered in post-its from the requirements workshop. Somewhere in all of that are the business objects she needs for the new data model. "Diego," she asks, "how long did it used to take you to figure out what actually needed to be modeled after a workshop like this?" Diego smiles. "Back in the day? Sometimes a week."
AI systems require perfectly structured data but cannot create the necessary data models themselves. Why does even the most powerful AI fail to understand what a "customer" or "product" means in a specific company? And why is precisely this definition work the key to success for every AI implementation?
Artificial intelligence is currently revolutionizing virtually every business area. Yet amid all the enthusiasm for these technologies, a fundamental paradox is often overlooked: AI requires high-quality, structured data to function at all. At the same time, AI itself is unable to create the data structures it needs to work.
How can I use natural language in my modeling process to achieve high-quality information models?
Resilient and temporally correct dimension Id’s for fact as well as dimension tables
In a dimensional data model based on a Data Vault data model, are integer values still necessary as dimension Id’s? And if so, how can the dimension Id’s be provided correctly?
Page 1 of 5