Enterprise AI fails at the operating model.
I help enterprise leaders turn AI from a set of pilots into an operating capability. My focus is the layer most programmes miss: ownership, workflows, decision rights, controls and the partner ecosystem around them. I bring more than 20 years across enterprise software, cloud, data and AI, including 12 years inside systems integrators and current UK&I partner leadership at Alteryx.
Ideas in context
From the operating-model thesis to the work it changes.
Operating model
- The Frontier Is Jagged. Your Workflow Needn't Be.
- The Most Dangerous AI Is the One That Does Exactly What We Asked
- Done, Keys, and a Second Check
- From 100 Agents to Two Heroes
- What Karpathy Didn't Say About Your Enterprise AI Strategy - Part 1: The Diagnostic
- Your Data Is Not Your Moat. Your Learning Loop Is.
- Tools Are Not the Operating Model
Agents
Governance
Adoption
Other ideas
Latest
The Trust Layer for Enterprise Agents Sits Outside the Model
Verify an agent's work outside the model, with deterministic checks where a rule can decide. A decision ledger and an independent checker set how far a client can safely scale agents.