Writing
How I think about AI, learning systems, community behavior, and marketplace design.
The AI features that fail are the ones trying to replace human judgment. The ones that succeed make humans better at being human. Here's the design framework I use to tell the difference.
You don't need to understand the math. You do need to understand what recommendation systems can and can't do — and how to design products around them. Here's the practical version.
The systems, the surprises, and the hard-won lessons from building a national digital learning ecosystem across 500+ SENAI schools in Brazil — and what it taught me about product at scale.
What the Model Context Protocol actually is, why it matters for product managers, and how to design an MCP integration that makes your product genuinely smarter — not just AI-washed.
How a simple weekly interview habit transformed product decisions at Voxy — and why Teresa Torres's continuous discovery framework is the closest thing to a PM superpower I've found.
Guardrail proposal for conversational discovery in a creative learning catalog: prioritize grounded retrieval, teacher-led content, and explicit fallbacks when confidence is low.
Prioritization memo for community work based on loop completion, not feature count. Focus is follow mechanics, feed distribution quality, and creator repeat contribution.
Operating memo for continuous discovery as weekly infrastructure: interview intake, synthesis, opportunity framing, and decision review.
Architecture memo for operating a national learning platform as a connected system across strategy, resource infrastructure, and learner-facing products.
Sequencing memo for expanding creator monetization formats without fragmenting discovery, profile, and trust flows.
Test comparing different follow and recirculation trigger sequences to improve loop completion in community flows.
Probe to test whether lightweight query rewriting improves retrieval quality for long, ambiguous learner prompts.
MCP architecture, LLM product design, semantic search, and AI-assisted decision support.
Creator economy mechanics, recommendation systems, subscription expansion, and post-acquisition integration.
Community feed design, retention loops, EdTech at scale, and discovery operating models.