Writing

Writing & arguments.

Exploring the intersection of AI, digital experience platforms, and software architecture. Insights from building and scaling enterprise systems in the age of AI.

  1. TechGuilds

    AI, You Lie!

    Draws on eight months leading AI at Kajoo Agentic to argue that building reliable agentic systems is less about model capability and more about the architecture around the model. Uses a real verification failure — an agent that passed a blank page because it checked HTML instead of the rendered browser experience — to illustrate how tools, state, workflow gates, and precise definitions of "done" are what make agent autonomy trustworthy.

    • Traces a real failure where an agent verified a blank page as passing by using curl instead of a browser renderer, and explains how the fix required building a proper tool, blocking the wrong path, and turning the rejection into a teaching moment
    • Argues that context windows are not databases — state, workflow sequencing, and completion records must live in the platform, not in the model's context
    • Reframes the developer role as designing conditions under which agents can produce, verify, and be trusted — less about writing every artifact and more about knowing when to constrain, when to delegate, and what counts as done
    • Agentic AI
    • AI Architecture
    • Verification
    • Developer Role
    • Workflow Design
    • DXP
    Read article →
  2. CMS Critic

    Seeing the Agentic Forest Through the Trees: Andy Cohen on Taming the AI Wilderness

    Featured guest on The Critic's Corner podcast with host Matt Garrepy. A wide-ranging conversation tracing the path from a Commodore 64 kid and Mount Diablo park ranger to founding architect of Sitecore XM Cloud, and into today's work leading agentic platform development at Kajoo and TechGuilds. Covers the realities of the agentic enterprise, the risks of vibe coding, and why fundamentals matter more than ever. Listen on Apple Podcasts: https://podcasts.apple.com/us/podcast/seeing-the-agentic-forest-through-the-ai-trees/id1801794414?i=1000766962858

    • Why AI isn't displacing work, it's shifting and expanding it
    • Drawing the boundary between deterministic tools and probabilistic agents
    • Hybrid skills for a generation whose first instinct is to ask the model
    • Apple Podcasts episode: https://podcasts.apple.com/us/podcast/seeing-the-agentic-forest-through-the-ai-trees/id1801794414?i=1000766962858
    • Agentic AI
    • Kajoo Agentic
    • XM Cloud
    • Vibe Coding
    • Developer Experience
    • Podcast
    Read article →
  3. CMS Critic

    The Next Great DXP Will Sell Transformation, Not Software

    Feature wars are a dead end. The next wave of DXPs will win by changing how teams ship, learn, and govern. This article shows how to anchor platform strategy to measurable business transformation and cultural adoption, not just tools.

    • Defines a value chain for DXP that ties product, engineering, and change management
    • Illustrates a transformation roadmap with operating model shifts, not just feature rollouts
    • Outlines leadership metrics that prove impact across speed, quality, and margin
    • DXP
    • Product Strategy
    • Org Change
    • Operating Model
    • SaaS
    Read article →
  4. LinkedIn

    Lipstick on a Pig: Why Most 'Agents' Aren't Agents

    Many so called agents are thin scripts wearing AI makeup. Real agents reason, plan, and coordinate tools across changing contexts. This piece offers a litmus test for agentic credibility and explains why orchestration matters more than wrappers.

    • Agent credibility checklist across autonomy, memory, tool use, and recovery
    • Patterns for safe delegation and guardrails in production environments
    • How to separate marketing claims from operational reality
    • AI Agents
    • Orchestration
    • Safety
    • Production Readiness
    Read article →
  5. LinkedIn

    The Commits Don't Lie: XM Cloud's Future in Plain Sight

    Roadmaps whisper. Repos shout. By reading dependency shifts, module seams, and deployment scripts, you can forecast where a platform is headed. This article decodes XM Cloud commit signals to surface strategy early.

    • Commit signal taxonomy across naming, dependencies, and platform seams
    • How to distinguish refactor noise from roadmap intent
    • Practical tips for investors, partners, and customers reading public repos
    • SaaS Architecture
    • Engineering Signals
    • Roadmap Intelligence
    Read article →
  6. LinkedIn

    Twelve Lessons I'm Still Learning: A Letter to My Younger Self

    Leadership is a practice, not a promotion. These twelve lessons cover clarity under pressure, bias for delivery, and the discipline of kind feedback. It is a pocket guide for staying human while building hard things.

    • Trade careful for clear when stakes are rising
    • Build systems that make the right thing easy
    • Coach in verbs and numbers, not vibes
    • Leadership
    • Career
    • Execution
    • Culture
    Read article →
  7. LinkedIn

    The Future of Architects and Product Managers in the Age of AI

    Architecture is shifting from diagram stewardship to outcome ownership. As AI automates scaffolding, the winning architects act like product managers who shape bets, measure impact, and tune loops.

    • Outcome first design with experiment cadence and product telemetry
    • From reference diagrams to performance contracts and SLOs
    • Skills map for the architect to PM evolution
    • Architecture
    • Product Management
    • AI
    • SLOs
    Read article →
  8. CMS Critic

    Agent-to-Agent Personalization: Trust, First Impressions, and the Grudge Factor

    Agents will increasingly serve other agents. Personalization must account for machine to machine trust, first contact rituals, and long memory. This article proposes patterns that reduce nondeterminism and build cooperative behavior.

    • First impression protocols for agent handshakes
    • Managing memory, reputation, and forgiveness across agents
    • Designing for graceful failure and negotiated retries
    • Personalization
    • AI Agents
    • Trust
    • Governance
    Read article →
  9. LinkedIn

    Building to Standards That Don't Exist Yet

    On the frontier there are no recipes. This article lays out how to build responsibly when patterns are immature, including how to design seams, isolate risk, and write standards that can survive version two.

    • Seam first architecture with well lit escape hatches
    • Risk isolation by domain, not by team
    • Drafting living standards that converge with use
    • Standards
    • Architecture
    • Governance
    • Experimentation
    Read article →
  10. LinkedIn

    The CMS is Dead. Long Live Structured Vibing.

    Rigid content models cannot keep up with adaptive experiences. Structured vibing pairs typed content with generative flexibility so teams can keep control while the system improvises within safe bounds.

    • Explains structure plus improvisation as a governance model
    • Shows how tokens and policies set creative lanes for AI
    • Maps use cases from content ops to on site orchestration
    • CMS
    • AI Content
    • Design Tokens
    • Governance
    Read article →
  11. LinkedIn

    Forget DevEx, The Future is AgEx

    Developers are no longer the only first class users of platforms. Agents are. AgEx focuses on giving agents clear tools, contracts, and feedback loops so they can work with us and for us at scale.

    • Designing APIs and events for agent consumption
    • Operational contracts for agent behavior and cost control
    • Metrics that track agent productivity and safety
    • DevEx
    • AgEx
    • APIs
    • Observability
    • Cost
    Read article →
  12. LinkedIn

    Accelerators Done Right: From Sales Demos to Real Delivery

    Most accelerators stop at the sizzle. This article details how to build accelerators that compress delivery time in production through reusable components, automation, and ops ready defaults.

    • Separates demo kits from delivery accelerators with entry and exit criteria
    • Defines reusable backbone assets across components, scaffolding, and CI
    • Links accelerator value to margin, quality, and ramp time
    • Accelerators
    • Delivery
    • SaaS
    • Reuse
    • CI/CD
    Read article →
  13. LinkedIn

    Three Lessons from Building and Scaling XM Cloud

    Behind the scenes of launching and scaling XM Cloud. Three lessons on aligning architecture with unit economics, turning continuous delivery into culture, and designing for cost aware growth.

    • Architecture choices tied directly to margin and tenancy economics
    • Culture of continuous delivery as an operating system for teams
    • Pragmatic patterns for scale without runaway complexity
    • SaaS
    • XM Cloud
    • Cost Optimization
    • Continuous Delivery
    • Tenancy
    Read article →

Topics

What I write about.

  • AI & Agents
  • Digital Experience Platforms
  • Platform Engineering
  • Software Architecture
  • Product Strategy