AI Leadership
Why Your AI Coding Agent Gets Expensive Before It Gets Useful
AI coding agents incur cost while searching, reasoning, editing, and retrying. Their value appears only when the resulting change is validated and accepted.
Writing
Technical and leadership writing
The writing below reflects my thought process, opinions, and evolving perspective on engineering, leadership, technology, and the decisions that shape resilient systems and effective teams.
AI Leadership
AI Can Broaden Contribution—But Who Maintains the Code?
Coding agents may expand who can contribute without expanding who can understand, review, and maintain the resulting code.
AI Leadership
The BlackBerry Trap of Enterprise AI
Why the winners in AI will redesign workflows, not just add agents to old ones.
AI Leadership
Human-in-the-Loop Is Not the Goal: How to Build Agents That Know When to Escalate
Enterprise agentic systems should not rely on humans for every decision. The stronger pattern is metacognitive controlled autonomy: agents that know when to act, verify, escalate, or stop.
Engineering
PoVSmith: Turning Vulnerability Findings Into Executable Evidence
A practical look at PoVSmith and how enterprise Java teams can move from scanner-driven vulnerability tickets to generated proof-of-vulnerability tests and remediation evidence.
AI Leadership
Explore-Then-Act: A Simple Design Pattern for Reliable AI Agents
A practical engineering-leader interpretation of Look Before You Leap: Autonomous Exploration for LLM Agents, explaining why reliable agents need a discovery phase before execution.
AI Leadership
Software 3.0 Needs Engineering Leaders
Software 3.0 is not simply AI writing code. It is a shift toward engineering systems around context, tools, memory, evals, security, and feedback loops.
AI Leadership
The Model Is Not Yours: What OpenAI's Goblins Teach Enterprises About AI Control
OpenAI's goblin incident is a useful case study in provider-side behavior drift, reward design, and why regulated enterprises need AI control planes at the workflow boundary.
AI Leadership
The AI Behavior Contract: How Enterprises Can Detect and Control Model Drift
A practical banking example showing how behavior contracts, evals, contract checkers, telemetry, and rollback thresholds can control LLM behavior drift.
Leadership
The Leadership Playbook for Sustainable AI Adoption
Why productivity gains are only the starting point — and why trust, work redesign, and talent development determine whether AI actually sticks.
Engineering
Operational Resilience Starts Before Production
A practical view of why resilience belongs in design reviews, testing strategy, canary releases, rollback readiness, dependency mapping, and team culture.
Engineering
TOON vs JSON: What They Are, How They Differ, and When to Use Each
A practical guide to JSON and TOON, with examples and a clear decision framework for AI-assisted development.
Leadership
What AI Changes for Engineering Leaders
A practical look at how judgment, communication, and engineering management evolve when software delivery speeds up.