Josh Wickes writes practical, first-person accounts of building AI agents and automations, including a Teams knowledge agent, an automated briefing, and an auditable Codex memory system. Readers get design decisions, failure modes and the reasoning behind them.

Short essays on how engineering decisions, vendor agreements, and automation hold up under pressure and over time. Readers get a reasoning framework for settling recurring questions, pricing exit costs, and judging when automation is truly finished.

Case studies of AI automation systems built and run in production by a single person, covering local-government meeting coverage, weekly show production pipelines, and cross-vendor telephony diagnostics. Each issue explains what the system does, its human-review guardrails, and measured results.

Praxis CTO publishes multi-part series on technology leadership, currently covering how to find and automate toil with AI and how to oversee AI agents using management frameworks. Readers get structured, practical methods for deciding where AI should be applied and how much autonomy to give it.