HWW/001 Web cover v0.4 2026-08
HWW/001 · The cover sheet

Human intelligence applied to the architecture of work.

Work, redesigned for people.

For a century, more of the capacity a job supplies meant more human hours. It doesn't anymore.

Most organizations are still designed as though it does. Ai can now perform, accelerate, or augment meaningful portions of productive work. Yet jobs, departments, headcount models, management systems, and the workday itself remain largely organized around human hours.

Human Workday studies what happens when the architecture catches up.

The different question

Ai capability is outrunning the organizations meant to use it.

Most organizations ask

How do we add Ai to the jobs we already have?

Human Workday asks

What work actually needs to exist now? And only then — who or what should perform it?

Every organization already has a Work Architecture™. It determines what work exists, how it is organized, where decisions sit, what capacity performs it, where accountability lives, and how work becomes coherent human roles. Some organizations design it deliberately. Most inherit it.

An organization without a Work Architecture still has one. It is simply unowned.

Work Architecture is the discipline of designing it deliberately.

The method

Don't start with the job. Start with the work.

The human boundary

Work gets measured. People don't.

Human Workday decomposes work, not human beings. It does not treat people as optimization variables, reduce human worth to productivity, or assume that everything technically automatable should be automated. The unit of analysis is the work. The object of design is the architecture. The human being remains a human being.

See it applied

Departments reimagined.

Theory is cheap. Architecture should produce something you can run.

People

Inherited jobs decomposed. Work allocated across human, Ai, system, hybrid, and external capacity. Human work recomposed. Roles redesigned.

Explore People

Information Technology

Same discipline. Different work. Different capacity. Different architecture.

Explore IT IN PREPARATION

Finance

Same discipline. Different work. Different capacity. Different architecture.

Explore Finance IN PREPARATION

Same method. Different answers.

The method doesn't prescribe the answer. Five inherited HR jobs produced materially different architectural conclusions:

HR Generalist Split and recompose People Partner
Recruiter Delegate and widen capability Talent & Capability Lead
HR Manager The thing being managed changes Head of People
HR Coordinator The inherited role may not persist People Operations Lead
Payroll & Benefits Administrator Much of the delegation is already system and external, not an Ai story Total Rewards Lead, where scale supports it

The method doesn't prescribe the answer. The work reveals it.

Working drafts. Destination titles are hypotheses to be tested against a work inventory, never recommendations — the method also preserves the inconvenient conclusions: keep the job unchanged, use fractional capacity, externalize the work, or find that a standalone role is not justified.

The human consequence

Architecture changes the workday.

Declared fictional — "Renata" is a construct for illustration; no real person or organization is depicted.

Before — HR Generalist

  • Policy questions. Nineteen more by noon.
  • I-9 corrections.
  • Enrollment problems.
  • Verification letters.
  • Manager chasing.
  • Interruptions.
  • The consequential employee investigation waits until 4:00.
  • Notes are being finished at 6:30.

After — People Partner

  • Systems handle routine questions.
  • Ai prepares and synthesizes.
  • Renata verifies the delegation.
  • Her time concentrates around credibility interviews, manager coaching, consequential interpretation, organizational patterns, and decisions requiring human judgment.
"The redesign did not remove the weight of the job. It removed the nineteen interruptions between her and doing it well."

That's the Human Workday.

Enterprise implications

Capacity before headcount.

Stop beginning with the org chart. The question isn't "how many people do we need?" until we've answered: "what work exists, and what capacity should perform it?"

TRADITIONAL GROWTH → HEADCOUNT → JOBS → TASKS

HUMAN WORKDAY REQUIRED OUTCOME → WORK → CAPACITY → HUMAN REQUIREMENT

Ai can inherit execution. It cannot inherit accountability.

As machine execution increases, accountability doesn't disappear. It moves.

Machine execution does not eliminate human accountability. It relocates it.

Rigor

We don't ask you to trust the diagram.

Every consequential architectural claim should be challengeable.

Evidence Grades How strong is the underlying evidence?
Delegation Grades How established is the proposed allocation?
Counter-cases What evidence argues against the architecture?
Durability Conditions When should the decision be reconsidered?
Architecture Decision Records Why was the work designed this way?

Inspect the research

Don't just read it. Run it.

Proof that the method produces Work Architecture a practitioner can run, not commentary about one. Worked examples, templates, checklists, job descriptions, architecture records, and implementation tools.

Don't copy our architecture. Run the method against your work.

The institution

An institute that happens to build software. Research precedes product.

  • Software may eventually instrument the discipline. It does not define the discipline.
  • Commercial convenience does not override evidence.
  • Product requirements do not override constitutional constraints.
  • Ai capability does not determine what humans ought to delegate.