eval-driven development
- AI Strategy (315)
- Generative AI (106)
- IT Leadership (33)
- Leadership (76)
- Product Management (344)
- Product Management Leadership (306)
- Uncategorized (32)
-
Reliable Agentic AI Architectures: A Production Blueprint
A production blueprint for agentic AI covering bounded graphs, independent verification, safe tool execution, durable state, and eval-driven rollout.
-
How to Run Customer-Facing AI Implementations That Scale
A practical operating model for turning vague customer AI requests into safe production deployments without creating a permanent custom-work queue.
-
Cost-Aware AI Model Selection: Pay for Accepted Work
A practical framework to compare AI models by accepted-result cost, test quality, choose hosting, and route production work with safe fallbacks.
-
How to Design Production AI Guardrails for Sensitive Workflows
A practical control-plane design for routing sensitive data, constraining agent actions, enforcing approvals, testing failures, and preserving auditability.
-
How to Build Trustworthy AI Diagnostics for Women’s Health
A practical framework for defining clinical scope, communicating uncertainty, preventing automation bias and governing sensitive health data.
-
How to Operate AI Customer Agents as a Reliable CX System
A practical operating model links agent capabilities, release readiness, evaluation, measurement, and ownership into one continuous CX improvement loop.
-
AI Product Leadership: Faster Learning, Safer Systems
A practical framework for improving product discovery and delivery while matching AI evaluation, privacy, and governance to real-world risk.
-
Reliable AI Coding Requires Four Kinds of Control
A practical framework for controlling requirements, context, verification, permissions, and recovery when teams build software with AI agents.
-
AI Inference Economics: Optimize for Value, Not Cost
A practical framework for balancing AI inference cost, latency, and quality against conversion, retention, support demand, and revenue.
-
How I Use Novus, the First Product Agent, to Turn Rapid Releases into Measurable Wins
Rapid releases don’t have to blur what’s working. I use Novus to connect code velocity to customer outcomes with eval-driven development, observability, and continuous discovery—without drowning…
-
Claude Code for Product Managers: Accelerate Prototypes, Validate Faster, Ship with Confidence
Claude Code helps me turn ambiguous product ideas into testable prototypes, grounded analyses, and instrumented experiments in a fraction of the time. By pairing it with…
-
Beyond Black‑Box Scores: Custom AI That Elevates Trust & Safety Without Burnout
Off-the-shelf moderation scores miss nuance and force teams into painful manual review. In this episode, I explore how Musubi blends traditional ML with LLMs, policy optimizers,…
-
How Agentic Analytics Reshapes Product Development Roadmaps
A practical synthesis of how product agents connect behavioral analytics, controlled experiments, roadmap decisions, and accountable automation.
-
AI Agent Product Development: From Workflow to Autonomy
A practical framework for developing AI agents through focused workflows, bounded capabilities, measurable outcomes, and evidence-based autonomy.
-
From AI Builder to Agent Swarm: A Product Delivery Model
A practical model for combining AI Builders, selective agent parallelism, evaluations, and release controls across product discovery and delivery.
-
Engineering MCP Agents as a Reliable Product Platform
A practical framework for building MCP agent platforms around controlled context, safe actions, measurable reliability, and governed scale.
-
Reusable AI Agent Workflows Need Evaluation Contracts
A practical model for packaging agent skills with traces, test fixtures, guardrails, and product metrics so reuse does not weaken accountability.
-
A Reliable Amplitude AI Workflow for Product Decisions
Build Amplitude AI analyses stakeholders can trust by controlling context, checking evidence, and turning findings into accountable product decisions.
-
How to Build a Resilient Experimentation Program at Scale
A practical operating model for producing trustworthy decisions through layered evaluation, governed measurement, and reversible delivery at scale.
-
An AI Operating Model That Measures Outcomes, Not Activity
A practical way to connect AI quality, customer behavior, revenue, risk, and delivery metrics so teams know what to scale, fix, or stop.
Weekly digest
One email a week on AI products, enablement and hiring. No fluff.
Browse topics
- AI Strategy (315)
- Generative AI (106)
- IT Leadership (33)
- Leadership (76)
- Product Management (344)
- Product Management Leadership (306)
- Uncategorized (32)
Work with me
45-minute consultation on AI product strategy, GTM and PM hiring — no charge.