Module 01
Why IE applies to AI
Capacity, variability, Theory of Constraints. The lens that changes how you see your team's work.
Free practitioner playbook — 100% yours to keep
Your AI team ships
on vibes.
This fixes that.
12 chapters applying industrial-engineering discipline — CPM, DOE, SPC, RACI — directly to AI engineering work. Practical. No framework fluff.
The methodology behind work at
Urban Group
V-Hola
Impilo
Datomize
TCM Digital
Who this is for
If any of these are true, this is the missing layer.
- Your team makes prompt changes based on "it felt better" rather than tracked eval data.
- You have no idea where the bottleneck in your pipeline is, so you optimize everything and see improvement nowhere.
- Your engagement timelines are someone's best guess — written on air, slipping in practice.
- You shipped an AI feature six months ago and genuinely don't know if it's still working properly.
Inside the playbook
12 chapters. Each maps a classical IE method to AI engineering work.
Module 02
Project plumbing
WBS, CPM, PERT, RACI, EVM. Scope and schedule any AI engagement without surprises.
Module 03
Methods engineering
Time studies and SOPs for AI work. Find the motion waste that costs your team 1–3 minutes every iteration.
Module 04
Operations research
LP for GPU scheduling. Queueing theory for latency planning. The math behind "which lever has the biggest effect."
Module 05
Quality engineering
SPC on agent metrics. DMAIC for prompt quality. Control plans that lock in gains.
Module 06
Simulation
Discrete-event simulation for AI pipelines. Monte Carlo for NPV under uncertainty.
Module 07
Engineering economy
NPV, IRR, and payback applied to deal pricing. The math that shifts you from cost-plus to value-based.
Module 08
DOE on prompts
Design of experiments turns prompt engineering from craft into science. 8 factors. One experiment. Statistically valid conclusions.
Module 09
Decision analysis
AHP for vendor selection, build-vs-buy, engagement prioritization. Decision trees for sequentially-dependent choices.
Module 10
Process mining
Discover the actual process from your event logs — not the one you think exists. Bottleneck analysis with PM4Py.
Module 11
Knowledge management
The KM system that survives engagements. Why most KM fails and how to make it live in the workflow.
Module 12
Capstone
A full-scope synthetic AI engagement, run end-to-end: WBS, CPM, DOE, SPC, NPV — all in one.
What's included
Everything you need to start tomorrow.
Full PDF playbook
Downloadable, yours to keep. CC-BY-NC-4.0 — share it, use it, cite it.
Companion GitHub notebooks
Runnable Python notebooks for CPM, DOE, SPC, DES, AHP, process mining.
Ready-to-use templates
WBS, RACI, control plan, DOE design sheet, Monte Carlo NPV — all in Excel.
Immediate access
PDF link in your inbox within minutes. No account, no paywall, no expiry.
What practitioners say
From the teams who ran it.
The DOE chapter alone changed how we structure prompt experiments. We went from "let's try a few things" to designed experiments with statistical conclusions.
I've read three "AI management" books and none of them gave me concrete tools. Chapter 2 on CPM/EVM gave me the scheduling model I actually needed.
The SPC chapter was what I was missing. We now run control charts on our agent's hallucination rate — we know immediately when something has drifted.
Already familiar with IE methods?
Get the AI Agent Engineering playbook instead.
Same discipline, applied to building, evaluating, securing, and shipping production agent systems.
Get the AI Agent Engineering playbook →