Blake DeBoer
Consulting Software Developer III
Born and raised in Chicago, Strider Alum Blake's blogs!
Posts by this author
Autonomous Coding Is Earned, Not Assumed
AI coding adoption fails for two reasons, and neither is the model. A three-part readiness test for deciding where autonomous agents can run, and where humans stay in the loop.
The Autonomous Coding Readiness Checklist
A three-part readiness test for engineering leaders: score your organization and every workstream to decide where AI agents can run unsupervised, and where humans stay in the loop.
Measure Fluency, Not Maturity
Maturity models rank you against a generic ladder. Why measuring code quality, delivery stability, and adoption together against your best teams is the real signal.
Individual AI Experimentation Doesn't Scale. Shared Rituals Do
Individual AI skill lifts individuals, not throughput. Why mob elaboration and mob construction turn one engineer's practice into how the whole team works.
Why Shorter AI Coding Sessions Produce Better Code
Model output degrades as context grows. Why bounding delivery scope and coding sessions with a plan-first discipline improves quality and cuts token cost.
Build a Knowledge Layer So Your AI Stops Relearning Your Business Every Session
Every AI coding session rebuilds context from scratch. A persistent knowledge layer of domain models and retrieval tooling encodes your intent so it doesn't.