DATA & ANALYTICS
“THE MODELS ARE ONLY AS GOOD AS THE CONTEXT THEY ARE GIVEN. FOR AN AGENT TO REASON RELIABLY, DATA NEEDS TO CARRY MEANING, NOT JUST VALUES”
Josh Fecteau Chief Data & AI Officer and Chief Information Officer Teradata
Q. HOW CAN YOU TURN ENTERPRISE DECISIONS INTO‘ TIME-BOUND SNAPSHOTS WITH FULL CONTEXTUAL METADATA’ WITHOUT CREATING MASSIVE OPERATIONAL FRICTION?
» The friction mostly comes from retrofitting context onto decisions that were never captured with it. The practical answer is to start at the moment a decision is made. But the idea of autonomy with AI is showing us that most companies aren’ t always capturing all the context that goes into making decisions. We rely on humans to do that.
Think about a pricing exception that gets approved for a specific customer during a rough patch in the relationship. The relationship recovers, the customer is back to full health, but the discount keeps applying automatically. An agent processes it every quarter for a reason nobody is looking for anymore.
The fix is unglamorous: assign an owner to every material business rule, not just the data but the reasoning behind it; build in a review clock; treat context as part of the data asset itself, not something that lives in someone’ s
128 October 2026