Writing
Blog.
Essays on the systems, frameworks, and human dynamics that make — or break — product delivery at scale. (scroll for more)
Get blog alerts in your inbox.
Free — no paid account required.
2026
10 minValuestream Episode 5: Context Is the Job. Why Reliable AI Agents Are a Context Problem, Not a Prompt Problem
The context-engineering episode of Valuestream. Reliable AI agents fail on context, not prompts: what the model can actually see when it acts. The context budget, the four failure modes, and how to run context as a delivery discipline with version control, evals, ownership, and observability.
8 minMembersDeprecation Debt: The Hidden Liability in AI-Accelerated Delivery
AI made writing code nearly free, but retiring it never got cheaper. That gap compounds into deprecation debt, a fast growing form of technical debt: dead paths, stale APIs, and un-migrated systems that quietly cap delivery throughput. Here is how to measure and govern it before it stalls your team.
8 minContext Engineering: The Delivery Discipline Behind Reliable AI Agents
Prompt engineering was never the real bottleneck. In production, agentic AI systems fail on context: what the model can actually see when it acts. Here is why context engineering is a delivery discipline, not a research problem, and how to run it with ownership, evals, and observability.
8 minMembersMean Time to Understand: Rethinking Incident Response for the Agentic Era
For a decade we optimized mean time to resolve. Now agentic AI detects and remediates in seconds, so the real constraint in incident response is comprehension. Here is why delivery teams need a new metric, mean time to understand, and an operating model built around it.
- 14 min
Valuestream Episode 4: The Demo Isn't the Deliverable. Crossing the Production Gap Between Pilot and Production
The production-gap episode of Valuestream. Most enterprise AI pilots don't fail on the model; they stall in the delivery and operating work between a dazzling demo and a system that survives production. The four layers of the gap: verification, evaluation, operability, and ownership, plus the one stalled pilot you can move this week.
8 minProgressive Delivery for the AI Era: Speed Without a Kill Switch Is Just Faster Chaos
AI made your team ship code faster; the 2025 DORA data shows it also made delivery less stable. Progressive delivery gives AI-era teams a control plane of canary releases, feature flags, and automated rollback wired to your error budget, so you ship fast without shipping chaos to production.
10 minMembersThe Merge Queue Is Your New Critical Path: Release Management When Agents Outpace Main
Agentic coding tools made writing code nearly free, but every change still has to pass through one serialized gate to reach production. This is a practical guide to treating the merge queue as the real critical path, sizing it, and keeping green main from becoming your delivery ceiling.
14 minThe Gentle Singularity: What Sam Altman's AGI Declaration Actually Means for Enterprise Leaders
Sam Altman declared we have crossed the event horizon into the singularity. Dario Amodei expects Nobel-caliber AI by 2027. $690 billion in AI infrastructure spending is committed for 2026 alone. Here is what the AGI declarations, the rhetoric shifts, and the capital flows actually mean for leaders building products and running delivery programs right now.