Home

🏠 Leverage Lab

Owner: Bill Eisenhauer Last Updated: August 7, 2026 Status: 🚧 Living research Contributors: Bill Eisenhauer

---

A practical guide to finding what slows software delivery when humans, agents, platforms, and partners all do part of the work.

Leverage Lab is a small, repeatable simulator. This wiki explains the thinking behind it and offers practices that can be used without copying the simulator's numbers.

New here? Start with Leverage Lab in Plain English. It explains the main idea and the important terms in about five minutes.

The argument in one picture

flowchart LR
    A["More build capacity"] --> B["More implementation output"]
    B --> C{"Can the rest of the system absorb it?"}
    C -->|No| D["WIP, rework, and attention pressure"]
    C -->|Yes| E["More accepted and owned outcomes"]
    D --> F["Locate the governing constraint"]
    F --> G["Run a bounded intervention"]
    G --> H["Collect a learning receipt"]
    H --> F

Loading

In prose: more building capacity helps only when the rest of the system can keep up. If it cannot, find the part that now sets the pace, improve it, measure the result, and look again.

Start here

If you want to…	Read…
Get the gist in about five minutes	Leverage Lab in Plain English
Understand the core claim	Operating Hypothesis
Understand why work should be ready before commitment	Kitting and Wait States
Design work across people and agents	Human-Agent Delivery System
Make the delivery system visible and choose an intervention	Flow Discovery and Intervention Rig
Run constraint-focused learning cycles	Constraint Learning Loop
Inspect the simulator's math and limits	Simulator Model
Use the lab with a team	Facilitator Guide
Trace the ideas to public material	Sources and Further Reading

What this is—and is not

This is a theory to test, not a benchmark, maturity score, transformation program, or diagnosis of a particular company. It combines established ideas about flow and constraints with a newer question: what changes when agents make implementation much easier to produce?

The reusable proposition is:

Measure value at the boundary where work becomes accepted and owned. Improve the current system constraint. Then look again, because a successful intervention changes the system.

Repositories

• Simulator source
• Interactive site

---

Next: Leverage Lab in Plain English