Series
Hands-on notes from building with AI agents: an app in a day, tests for an AI setup, a self-hosted personal assistant, and an agent that produces on-brand visuals.
The strategy posts elsewhere on this site argue for bounded scope, evidence and ownership. This series is where I actually build the things — first person, with the setup, the dead ends and the parts that turned out to matter more than expected.
Four projects: building a working iOS app in a day and finding the real speedup was in requirements, not typing; putting an AI setup under automated test instead of trusting a manual read-through; running my own assistant on a VPS so one agent has my tools, files and context; and wiring that agent to Figma so it produces editable, on-brand social visuals.
Part 1 of 4 ·
Discover how AI shifts the development process from mere coding to rapid requirements engineering and architectural sparring. This article explores a real-world experiment of building a functional, AI-powered iOS app.
Part 2 of 4 ·
Moving past basic chat interfaces to custom enterprise AI workflows requires scaling beyond simple intuition. This article explores why manual reviews fail to catch critical regressions when updating prompts, models, or agent behaviors. Learn how to treat your AI setups like core software and build operational confidence through automated testing.
Part 3 of 4 ·
Why I run my own assistant instead of relying only on provider apps.
Running Hermes Agent on a VPS gives me one assistant with my tools, files, context, recurring workflows, and review points across Telegram and email.
Part 4 of 4 ·
How I set up an AI agent with the Figma MCP and a component library to generate editable, on-brand social media visuals.
Regenerating AI images for every small tweak is slow and costly. How I generate editable social media visuals with an AI agent, the Figma MCP and a skill.