Notes
Occasional writing on applied AI, knowledge governance and B2B websites. No fluff.
Aug 23, 2026
Enterprise Context Engineering Platforms: Buy, Integrate or Build?
A practical guide to enterprise context engineering platforms, tools and services: what the category includes, how vendors differ, and when to buy, integrate or build your AI context layer.
ReadAug 19, 2026
Agentic RAG vs Traditional RAG: When Is It Worth It?
Agentic RAG plans, searches, evaluates evidence and can search again. That improves complex answers at the cost of latency, expense and new failure modes. This guide shows when traditional RAG is still enough.
ReadAug 19, 2026
AI Agent Governance: Permissions, Identity and Human Approval
AI agents need more than model guardrails. The seven controls every enterprise agent needs: identity, authentication, authorization, data boundaries, tool boundaries, human approval, and audit with a kill switch.
ReadAug 19, 2026
Is Your Knowledge Base Ready for AI? A 12-Point Checklist
Before deploying RAG or AI agents, check whether your knowledge is accurate, current, owned, structured, permissioned and traceable. Score yourself against this 12-point checklist and see what to fix first.
ReadAug 19, 2026
Forward Deployed Engineer vs Solutions Engineer vs Consultant
An FDE builds and owns production systems inside your business; a solutions engineer supports the sale of one product; a consultant diagnoses and recommends. Here is which one your company actually needs, phase by phase.
ReadAug 19, 2026
Hiring a Forward Deployed Engineer: Full-Time vs Fractional
Should you hire a full-time FDE or use a fractional model? An honest comparison of cost, ownership, engagement length and scope — including the situations where you should not hire an FDE at all.
ReadAug 19, 2026
Context Engineering for Enterprise AI: Why Prompts Aren't Enough
Context engineering is the discipline of designing what an AI system can know, see, remember, retrieve and act on. This guide explains the seven-layer enterprise stack and why better prompts cannot fix a broken context architecture.
ReadAug 17, 2026
AI Customer Support That Cites Sources Beats a Chatbot That Guesses
Most AI support fails on trust, not technology. How source citations, honest escalation and clean knowledge turn a chatbot into a front desk that wins orders at 2am.
ReadAug 15, 2026
Knowledge Governance: the Most Important, Most Exhausting, Dirtiest Job in Your AI Plan
Why enterprise AI projects fail on messy knowledge, not weak models. What knowledge governance actually involves, with a worked example, a pipeline diagram, and a pragmatic 90-day start.
Read