What this course is
One method to identify, prioritize, design, architect, build, test and deploy automation, AI and agentic solutions that deliver measurable outcomes.
One method every functional team can use — finance, human resources, procurement, customer operations, risk, technology — to take an idea from “AI could help here” to something running in production with controls that hold. It is vendor-agnostic on purpose: the method, the vocabulary and the artifacts are stable, and the platform specifics belong in the specialist courses that follow it.
8 hours
Nine modules, seven artifacts, one worked case carried end to end.
7 rungs
From rules and automation to multi-agent systems — and the reason most candidates should stop well before the top.
12 factors
Weighted value and complexity scoring on written anchors, so a portfolio ranks instead of averaging to 3.5.
The spine
Five phases. The questions are the deliverable; the letters are how you remember them.
A — Align
What outcome matters, what is the baseline, and why is this the smallest sufficient solution?
G — Ground
How should the work itself change, who decides what, and what evidence does the system need?
E — Equip
What may it know, decide and call? Instructions, context, tools, skills, schemas, loops.
N — Narrow
What must be prevented, approved, detected, reversed or escalated — outside the model?
T — Trace & Tune
Is it reliable, observable, affordable and supportable? Evaluations, release, operate.
The modules
Each one states what a participant can do at the end that they could not do at the start.
0. Opening
20 min
Know why the method matters more than the model, and what may and may not be brought into a shared learning environment.
Why method beats model · Working with real material safely · The five-phase spine · Eight delivery doctrines
1. The solution landscape
40 min
Place any candidate on a seven-rung ladder and say out loud why the cheaper rung below it will not do.
Rules and automation · Predictive ML versus GenAI · Single calls and retrieval · Workflows versus agents · Multi-agent systems and what they cost
2. Identify and prioritize
57 min
Turn “AI could help everywhere” into three things to do first, a sequence, and a written list of what is being declined — defensible in front of a steering committee.
Where candidates come from · The five gating questions · Value and complexity scoring on written anchors · The four quadrants and four operating models · Sequencing and declared declines
3. Deconstruct and reimagine
62 min
Take an end-to-end process apart to activity level and redesign it — rather than automating what is done today.
Activity mapping with touch and elapsed time · Eliminate, simplify, standardize, determinize, delegate · The authority matrix · Autonomy tiers and promotion criteria
4. From prompt to harness
47 min
Move from writing better prompts to engineering the whole context and then the whole harness the model runs inside.
Prompt engineering and structured output · Context management: what goes in the window and what does not · Context engineering: retrieval, compaction, state · Harness engineering: the model is one component
5. Architect and build the harness
75 min
Compose a solution from named parts — skills, tools, schemas, loops, hooks, orchestration — and write the contracts down.
Skills and tool design · JSON schemas and structured outputs · Loops, bounds and termination · Hooks as controls, not instructions · Orchestration patterns and multi-agent coordination · Tool integration and interoperability protocols
6. Prove it
66 min
Define success and severe failure before tuning anything, then build the evaluation suite that gates the release.
The five evaluation case types · The grader pyramid: assertions, schema, calibrated judge, human review · Confidence calibration · Human review design and provenance · Validation, retry and feedback loops
7. Deploy and operate
46 min
Ship it behind a release gate, watch the right signals, and manage cost per accepted outcome rather than token price.
Batch processing strategies · CI/CD integration and the evaluation gate · Tracing and observability · The release gate and staged autonomy · Unit economics and cost per accepted outcome · Incident runbooks and reversal
8. Close
17 min
Leave with the seven artifacts, an assessment route, and a named next step.
The seven-artifact package · Assessment and the hard fails · Specialist paths · Commitments
Who it is for
Practitioners who will scope, design, build or govern AI in a functional process — and the people who have to defend those decisions to a steering committee, a risk function or an auditor. No coding is required to complete it; the labs are design work, not implementation. It is the baseline for the specialist tracks that go deep on a specific vendor's agent SDK.
If you are new to agents altogether, Agent Building Foundations is the free, ungated place to start.
The full course is access-controlled
Everything above is the public overview. The modules, the scoring anchors, the labs and the reference material open with an access code. One code, and it lasts 24 hours in this browser.
No code? The course is run as a facilitated program — get in touch and we will point you at the next cohort.