Getting started with OpenProse: a small workflow you can inspect
A beginner OpenProse walkthrough: two Markdown files, one agent run, and how to check the saved result against its own requirements.
Insights & Knowledge
I transform complex technical ideas into scalable learning materials, talks, and documentation.
Core Themes
AI Memory & MCP
Backend Patterns
Human-Centered AI
Dev Education
Rapid Prototyping
Latest Articles
October 27, 2026
October 11, 2026
Agents Are More Than Just Skills
Sharing a colleague's method is useful. Collaborating with the agent they already work with preserves a different kind of continuity.
December 21, 2024
Full-Cycle Product Engineering: Shipping the Official Xano Node for n8n
From opportunity identification to community adoption—how I owned the complete product lifecycle of an ecosystem integration that unlocked 400+ automations for thousands of users.
December 20, 2024
Building a Self-Improving AI: The Evaluator-Optimizer Pattern
One-shot LLM calls aren't enough. Here's how I built a translation agent that critiques and refines its own output—achieving 40-60% quality improvement through iterative refinement.
December 19, 2024
Progressive Disclosure: How AI Agents Self-Corrected 7+ Errors Without Me
Bridging Language Server Protocol with Claude Code via MCP enabled autonomous syntax validation. Here's how progressive disclosure makes AI agents smarter, not just faster.
December 18, 2024
How I Built Cross-Platform AI Memory in 3 Days
Most AI problems are actually backend problems. Here's how I shipped SessionBridge—a memory layer that works across Claude, ChatGPT, and Cursor—using MCP and Xano.
December 12, 2024
Why AI Memory Matters More Than Context Windows
Long context windows are impressive, but true AI assistants need memory systems that compound over time. Here's why and how to build them.
Speaking & Workshops
Request a Talk Workshop
Building Agent-Ready Backends with MCP
A hands-on workshop teaching builders how to design backends specifically for AI agents using Model Context Protocol.
Conference Talk
AI Memory Systems: Making Every Interaction Count
Exploring how to build memory-first AI systems that compound value over time rather than resetting context.
Panel
Human-Centered AI: Beyond the Hype
Designing AI experiences that users trust, focusing on transparency, control, and progressive disclosure.
Talk
From 0 to Production: Shipping AI Fast
Rapid prototyping methodologies for shipping AI products without sacrificing quality or maintainability.