› Writing
Front-door takes on Microsoft, Anthropic, and the agentic runtime.
Long enough to be useful, short enough to read in one sitting. Written by people who’ve been inside the tenants and the rollouts these pieces are about.
Vendor reality · Coexistence
Where does Anthropic fit, as a newcomer?
Anthropic Claude Enterprise is not replacing Microsoft 365 Copilot — and pretending it is wastes everyone's time. Realistic look at where Claude wins inside enterprises that already run on Microsoft, where Microsoft Copilot wins, and how to deploy both without fragmenting your governance posture or your procurement leverage.
Read →Architecture · Walkthrough
Have you considered containers for runtime?
Most enterprise agentic stacks are monoliths in disguise — single Lambdas, single notebooks, single managed sessions you cannot operate, version, or swap a model out of. Containers are not the only answer, but they are the answer most production agentic systems converge on. Here is what the runtime decision actually looks like.
Read →Licensing · The new tier
Managed agents are here. Where are you?
Microsoft has shipped agentic workflows under a new licensing model. Most enterprises are being asked to decide on agents before they have the architecture to run them. The procurement conversation has to come first — and most teams are walking into it underprepared.
Read →Governance · Sequencing
AI is coming, ready or not. What's your Purview posture?
The day Microsoft 365 Copilot turns on, every 'anyone with the link' SharePoint share becomes searchable AI surface. Sensitivity labels, DLP enforcement, and the oversharing audit are not post-launch hygiene — they are the gating step. Most stalled Copilot rollouts stalled here. Here is the sequence that unblocks it.
Read →