Essays on keeping AI systems working in production. https://iggym.github.io/ 2026-06-01T00:00:00Z Iggy Mwangi https://iggym.github.io/articles/context-orchestration.html 2026-06-01T00:00:00Z 2026-06-01T00:00:00Z

For over a quarter of a century, software engineering treated code syntax as its primary unit of leverage. As production systems transition to generative runtime engines, the high-leverage skill moves entirely into structural prompt composition, token safety boundaries, and validation pipeline tooling. (6 min read)

https://iggym.github.io/articles/token-economics.html 2026-05-01T00:00:00Z 2026-05-01T00:00:00Z

An architectural review analyzing how long-tail conversation data, multi-agent overhead parameters, and strict JSON output schema rules quickly impact inference latency budgets inside production enterprise ecosystems. (4 min read)