Writing
Practical thoughts on software design, APIs, cloud systems, developer tooling, and the occasional experiment.
Featured article
In integration-heavy EV platforms, protocol diagnostics are essential for understanding what was sent, what came back, and how responses were interpreted. This article walks through improving a best-effort OCPI diagnostics pipeline by making it non-blocking, bounded, batched, observable, and safer under traffic spikes.
Read article →From the archive
distributed-systems
In integration-heavy EV platforms, protocol diagnostics are essential for understanding what was sent, what came back, and how responses were interpreted. This article walks through improving a best-effort OCPI diagnostics pipeline by making it non-blocking, bounded, batched, observable, and safer under traffic spikes.
10 Aug 2026
distributed-systems
Many engineers treat HTTP 429 responses as errors, but in modern distributed systems they are often a sign that protective mechanisms are working correctly. This article explores rate limiting, throttling, retry storms, cascading failures, and how healthy platforms use 429 responses to preserve stability and availability.
10 Jun 2026
projects
By using Catdocs, each team can manage their part of the API independently—paths, schemas, components—while still contributing to a single source of truth that is validated and bundled in CI.
19 Jun 2024