EnterpriseDataEngineering

The board wants agentic AI progress by Q3. The SI sends a team. Fourteen weeks later, nobody has looked at the actual data. There’s a different model: 90 days, a small team with deep domain expertise, and a working proof of concept against real production data. This article describes what an engineering-led agentic readiness sprint looks like — and why the traditional SI model isn’t built for this moment. Final article in the #EnterpriseDataEngineering series.

McKinsey called ERP “the ugly stepchild” of the AI conversation. This article maps the architecture of what agentic-ready actually looks like — from event-driven integration surfaces and structured business event streams to prevention-first data governance. Includes a detailed breakdown of application logs vs business events and why the difference determines whether your agent reasons or hallucinates. Article 2 of the #EnterpriseDataEngineering series.

42% of AI initiatives failed in 2025. Not because of the models — because of the data infrastructure beneath them. Three layers sit between an AI agent and a working enterprise deployment: APIs that can’t keep pace, business event streams that don’t exist, and master data that nobody governs. The first article in the #EnterpriseDataEngineering series.