ThinkingReal-world experience from the trenches
Practical insights on AI implementation, product strategy, and building systems that scale. 25+ years of senior consulting distilled into actionable guidance.
Latest Article
rks for Business Process Management — AI First, and the Interface to Your Business's Brain
AI First business process management: rks runs a defined process through a governed agent, grounded in your business's own local knowledge graph — its brain — and acting on your systems over MCP under plan → gate → record. The governance, audit trail, and cost visibility that make it trustworthy for software generalize to any defined process. You can connect business apps over MCP today (with manual setup); the turnkey, skills-complete product is still ahead. Honest about capable-today vs. still-ahead throughout.
More Articles
rks for Product and Design — Governed Thinking, Not Just Governed Code
A value-first tour of rks for product managers, designers, and design leaders: the PO Governor as a real product-owner step, concern-based scoping, a visible story pipeline, traceable rationale, first-class research (internal and external), cost visibility, working design tooling (Figma via MCP, browser-driven UCD passes), and the shift from directing tools to reviewing outcomes. Grounded, with vision clearly labeled.
RouteKit Shell — The Current Architecture
A present-tense overview of RouteKit Shell (rks) today: the Dispatcher → Governor → Agent tiers, the skill vocabulary, the story lifecycle state machine, concern-based decomposition, redirect hooks, autonomous CI and cost observability, RAG grounding, and the multi-project registry. Breadth-first, with per-feature deep dives to follow.
RouteKit Shell Agentified Workflow Deep Dive
Technical deep dive on how RKS evolved from tool-level orchestration to a multi-tier agent architecture. Covers the Dispatcher/Governor/Agent hierarchy, the Product Owner → QA → Build pipeline, context window conservation, and observations from production UAT runs.
RouteKit Shell Workflow Deep Dive
Technical deep dive explaining how RKS orchestrates AI agent workflows with guardrails, hooks, governance, and telemetry. Covers the core Detect → Plan → Validate → Apply pattern applicable to code, business automation, and beyond.
06 - The Learning System: How Contextual Intelligence Gets Smarter Over Time
Part 6: Breaking down how contextual AI systems continuously improve from every interaction, decision, and outcome
05 - Contextual Intelligence Beyond Engineering
How small businesses can leverage AI agents across marketing, finance, customer support, and operations
01 - Why Vibe Coding Fails
Getting LLMs to write JavaScript is trivial - the hard part is getting them to think and execute in valuable ways
04 - from Architecture to Product Intelligence
How solid AI-driven technical foundations unlock intelligent product management and design research
03 - Agent Decision Trees: How Contextual Intelligence Actually Thinks
Part 2: Breaking down the decision logic that transforms retrieved context into intelligent actions
02 - Anatomy of a Retrieval: How Smart Routing Actually Works
Part 1: Technical breakdown of how the Guardrailed Retriever Stack actually works
Atomic Documentation Architecture: Optimizing RAG Systems with Knowledge Graph Atomic Units
We restructured our documentation into atomic units and achieved 2x better RAG query performance, perfect content reuse, and dramatically improved maintainability. Here's how atomic documentation architecture transforms both human and AI workflows.
UX287: Principles for User-Centered Problem Solving
Exploring UX laws through the UX287 lens: actionable principles for impactful products driven by research, strategy, and collaboration.
Stay Updated
Get practical AI insights and RouteKit updates delivered to your inbox. No spam, just actionable guidance from the field.
Get Updates