Post

Over the Christmas holidays, I built three apps to replace my Google Sheets workflows: - 𝗙𝗶𝗻𝘁𝗿𝗮𝗰𝗸: A budget tracker that imports bank transactions and allows us to categorize them - 𝗣𝘂𝗺𝗽𝗲𝗹: A gym tracker - 𝗛𝗮𝗯𝗶𝘁𝘀: A daily habit tracker

All three started as spreadsheets I'd been using for months or years. Now they have proper UIs and additional features. Why the upgrade? AI made it possible 🥰 But here's what I learned:

I approached each project with different levels of "vibe coding." For Fintrack, I wrote most code myself or carefully reviewed every line. I wanted to understand exactly where data is stored and how it flows through the system (for obvious reasons 😅).

For Pumpel, I ran a pure conversation experiment: How far could I get without even looking at the code? And yes, it works. But I don't like it. Worse, adding new features through prompting gets progressively harder.

Fintrack, on the other hand, is a joy to work with. It has a robust foundation and architecture that makes it easy for both me AND the AI agent to know where features belong and how to reuse existing logic.

The lesson: AI coding tools work best when you establish clear structure upfront. The time you invest in architecture pays dividends in maintainability and makes AI assistance more effective, not less.