
What PMtivity is teaching me about building with a small team
Building PMtivity is reinforcing a simple lesson: PM leverage in 2026 comes from narrow workflows, visible judgment, and shipping useful artifacts faster.
Lessons from building products at scale. No fluff, just what actually worked.

Building PMtivity is reinforcing a simple lesson: PM leverage in 2026 comes from narrow workflows, visible judgment, and shipping useful artifacts faster.

Side projects usually do not die because the idea is weak. They die because the workflow ignores your real constraints, energy, and life season.

A good weekly PM operating system turns AI tools, prototypes, and recurring updates into a repeatable cadence that protects judgment instead of adding workflow noise.

Most AI agent failures are not model failures. They are approval failures. Here is where human checkpoints belong and how to design them without killing speed.

AI teams stop being demos when prompt, model, tool, and workflow changes hit evals in CI/CD before release. Here is what that release gate should actually cover.

AI made builders faster, but slow approval loops, bloated handoffs, and manager-heavy orgs are becoming the real bottleneck. Here is what to change first.

AI search is splitting traffic from citations. Homepages catch more clicks, while deeper comparison, reference, and evaluation pages earn more citations.

The rise of the forward-deployed engineer is the clearest signal that AI value has moved from model demos to real deployments. Here is what the role actually means, why it is spiking now, and what smart teams should do with that signal.

AI workflows do not usually break all at once. They decay through overlapping tools, stale instructions, polluted memory, and missing review gates. Here is what agent debt looks like in practice and how to avoid it.

A non-engineer PM's guide to shipping real products with Claude Code—from iOS apps to automation systems.

Practical tactics for Answer Engine Optimization—making your content discoverable by ChatGPT, Perplexity, Google AI Overviews, and Claude.

How a product manager with no Swift experience shipped an iOS finance app using Claude Code, SwiftUI, and TCA.

How I collected patterns from hundreds of prototypes to teach AI coding tools what makes designs feel generic—and how to avoid it.

How we built 50+ tools in one year with 2 engineers and scaled to 10 million monthly users in 18 months.

Building AI-powered workflows that transformed our organic acquisition capabilities at Picsart.

The systems thinking approach that let us ship 2 full flows per week with minimal resources.

A tactical breakdown of the changes that dramatically improved our landing page performance—plus how we automated LP creation afterward.

Reflections on shipping AI-generated audio news at NU.nl before the current AI boom, and what being early taught me about AI adoption.