The codebase was the brief for AI agents
I built the frontend foundations of an order management system. Agents completed much of the remaining implementation in days; the later pull requests needed mostly minor corrections.
Software engineer / Medical doctor
AI in the process. My name on the result.
A few things I've madeI built the frontend foundations of an order management system. Agents completed much of the remaining implementation in days; the later pull requests needed mostly minor corrections.
I built the review interface and Word integration for a contract tool. An approved correction had to reach the right passage without disturbing the surrounding revisions.
I built and maintain a medical exam platform with 3,000+ registrations, without paid advertising. The product question behind offline support was which progress a learner could trust.
I built React Overlay Manager so application logic can continue while a dialog finishes disappearing. That small distinction shaped the library's lifecycle and public API.
I implemented GA4 and Google Ads tracking for Lumo Elektro and repaired a silent delivery failure. Then I defined what the owner could actually conclude from the reports.
I resolved around 300 frontend findings and delivered ahead of the Google audit deadline. Later modernization cut build memory from 3.7 GB to 1.25 GB.
Building a product or a module that agents will help ship? Describe one case ↗