Skip to content

Lift 1: Gear Check — The Disruptor's Toolkit

Why a Gear Check?

You check your gear before you drop into a line — not because you don't know how to ski, but because starting on the same page matters more than starting fast.

This lift is a rapid calibration of the foundational concepts from the Black Diamond track: continuous delivery, evaluation harnesses, structured logging, quality gates, autonomy decisions, and orchestration patterns. You've likely been operating at this level already. If some of it is new, lean on your teammates — the team is only as strong as its shared vocabulary. If you already know it, you have a role too: help your teammates get there. Teaching sharpens your own thinking.

The goal isn't to test anyone. It's to ensure your team shares a common foundation — same terminology, same frameworks, same mental models — so that when Lift 2 introduces the shared stack and multi-contributor coordination, you're all operating from the same starting point.

Deployment in Coder is new for everyone regardless of experience.

What You'll Learn

  • What the pre-configured deployment pipeline gives you and why deploying before building matters — even when you already know the pattern
  • How golden datasets, evaluation harnesses, and structured logging create the measurement infrastructure that makes autonomy safe
  • How quality gates, orchestration patterns, and the autonomy slider let you design systems that direct themselves — and where humans MUST stay in the loop

Sections

  1. Deployment in Coder — What you get for free and why you deploy before you build
  2. Evals & Observability — Golden datasets, evaluation harnesses, and structured logging — the infrastructure that makes measurement-driven development possible
  3. Directing AI Systems — Quality gates, orchestration patterns, and the autonomy slider — designing systems that direct themselves

By the End of This Lift

  • Your application is deployed to a live URL before you write your first feature
  • You can describe the difference between tests and evals — and why correctness requires both
  • You know how golden datasets and structured logging create a closed-loop feedback system
  • You can apply the autonomy slider and reversibility principle to determine where humans stay in the loop
  • You know at least two orchestration patterns for critical decisions
  • Your team shares a common vocabulary: golden datasets, eval harness, trace IDs, three levers, quality gates, ratchet effect, autonomy slider, reversibility principle, delegation contracts