Stop Overbuilding Your Data Stack Before You Start — Mark Hay

Mark Hay, CTO of TextQL, in a 27-second TwoSetAI excerpt on planning data and AI work.

Don’t make the foundation the whole project

• Hay sees teams insist on consolidating all data, making curated “gold” tables for every business area, and defining every column in a semantic layer before adding capabilities.
• He says the more common mistake is overbuilding up front, not starting with too little infrastructure.
• The clip argues for questioning that sequence; it does not specify which minimum data controls to keep or demonstrate a replacement architecture.

“I much more frequently see people erring on the side of overbuild the solution.” — Mark Hay