Reliable automation
Idempotent jobs, reconciliation, explicit state machines, observability, and recovery after partial failures.
About the author
I'm johnnyji, a software engineer based in Seoul, South Korea. My work sits around web applications, crypto infrastructure, trading automation, and the operational tooling needed to keep automated systems understandable.
I started johnnyji.dev as a place for practical calculators. While building them, I realized the useful part is rarely the formula alone. Developers and operators also need to know which assumptions can fail, what data should be recorded, and how a result changes when fees, retries, stale data, or provider limits appear.
The articles on this site document those details. They are written from implementation experience and focus on decisions that can be tested: how to prevent duplicate orders, reconcile remote state, estimate API costs from product behavior, and place limits around systems that spend money automatically.
Idempotent jobs, reconciliation, explicit state machines, observability, and recovery after partial failures.
Execution costs, position risk, exchange integrations, inventory tracking, and the operational controls around strategy code.
API usage, infrastructure budgets, free-tier cliffs, and keeping experiments affordable until demand is proven.
The trading material is educational and does not provide financial or investment advice. Questions, corrections, and technical discussion are welcome through the contact page.