Shared state
under load.
Rate limiting becomes interesting when several application instances must agree on the same truth. The experiment focuses on atomicity, refill semantics, retry behavior and what Redis changes about the design.
04 / LAB · QUESTIONS WORTH EXPLORING
A technical shelf for the ideas that keep appearing across my builds — concurrency, graphs, service boundaries and algorithms.
Rate limiting becomes interesting when several application instances must agree on the same truth. The experiment focuses on atomicity, refill semantics, retry behavior and what Redis changes about the design.
Instead of treating transactions as isolated rows, model them as relationships. Connected components, cycles and anomaly scoring create a different lens for suspicious behavior.
Service decomposition is useful only when the boundaries create clarity. I explore how authentication, account state and payments can remain independently understandable while still composing into one product.
ALGORITHM PRACTICE
Problem solving and constraint thinking stay in the loop because good system design still depends on fundamentals.