Products and systems built around real constraints, from reliable automation to multi-tenant applications.
I built a local-first radar to reduce the time between a job being posted and being able to act on it.
The system combines continuous collection, deduplication, fit classification, and resilient Telegram alerts while keeping the main processing on the local machine.
I turned a discovery step based on scattered conversations and documents into a structured response flow.
People answer from their phones without creating an account; the Worker validates the data, stores private responses in D1, and sends an operational summary through Telegram.
I built a visual workspace that turns ideas into editable, persistent maps.
The same board combines mind mapping and a freeform canvas, with full backups, local persistence, and support for continuing work offline.
I built a complete operating system for a barbershop, connecting appointments, inventory, and finance.
The multi-tenant architecture and offline support were designed to keep the system useful at the counter even when connectivity fails.
I built a multi-tenant platform that connects reservations, operations, and financial visibility for sports arenas.
The product brings public bookings and administrative routines into one foundation, reducing the distance between customers and daily operations.
I'm building a point-of-sale system to reduce friction at the counter.
The system connects payments, sales, inventory, customers, and alerts in a simple operation for a local market, focused on real use by a small team.