Domain 4 – Workflow Integration and Solution Design (16%)

• Use Claude for requirements analysis and use-case exploration
• Support research, planning, and process optimisation
• Integrate Claude into existing business workflows
• Communicate Claude’s value and limitations to stakeholders

CCAO-F Domain 4 · Task 4.1

What You Need to Know

Claude structures; the human validates and owns the result.

Structure from Mess

Real requirements arrive as complaint threads, meeting notes, survey comments. Claude clusters them into themes, separates what people asked for from what they appear to need, surfaces incompatible demands, and names what nobody mentioned.

A Theme Is a Hypothesis

A theme Claude names is a hypothesis about your input, not a finding about your business. The people whose requirements it describes must confirm it before anything gets designed on top.

The Shape of a Strong Candidate

• Language-shaped work (drafting, summarising, reformatting, explaining, comparing)
• Material needed is material you can provide
• Output can be checked by someone competent to check it
• High volume, or the blank page is the expensive part

Weak candidates need authority to decide, depend on information Claude has no access to, or are relationship work rather than document work.

The Checker Test

If you couldn’t tell whether the answer was wrong, you’re not ready to use Claude on that task. Without a checker, a wrong output passes as easily as a right one. High stakes call for a competent checker in the process, not abandoning the drafting help.

Key Concept: When Claude’s freshly clustered themes are on the table, take them back to the people who own the requirements before design begins.

Exam Traps

• Adopting Claude’s themes as validated findings because they arrived quickly and read plausibly
• Handing Claude the decision along with the analysis
• Ruling work out as too important for Claude to touch

Practice Scenario

Saturday street market manager has a folder of complaint emails and trader comments about pitch allocation. Claude clusters into six themes, including a contradiction between regulars and casuals about corner pitches. She has no time to re-read the folder. What should she do next?

Build Exercise: Cluster, Confirm, Then Classify (30 min)

CCAO-F Domain 4 · Task 4.2

What You Need to Know

Claude produces the first pass; you own the judgement that follows.

Research at Two Scales

• Web search — straightforward factual queries answered in one or two lookups
• Research mode — multiple searches that build on each other; cited findings; bigger multi-source jobs

A citation shows where a finding came from, not that it is right. Check anything important before repeating it.

Plans Are Built from Stated Constraints

Claude plans confidently around constraints you never mentioned. Nothing in the output warns you. Put the real constraints in the brief, then ask for the assumptions it rests on. An unconstrained plan that reads like competence is more dangerous than an obviously bad one.

The Optimisation Loop

1. Describe the process and its constraints
2. Ask for candidate bottlenecks, and the evidence that would confirm each one
3. Go and get that evidence
4. Change the step the evidence indicts

The most expensive mistake is optimising a step that was never the constraint. Plausible and real causes coincide often enough to tempt you, and rarely enough to be expensive.

Key Concept: In this objective’s stems, the keyed answer checks the evidence before changing anything.

Exam Traps

• Treating a plausible, well-structured plan as a checked one
• Changing the process on the strength of the best-sounding hypothesis
• Substituting more research for local evidence

Practice Scenario

City bike-hire scheme coordinator hears that commuter-belt docking stations are empty by 8am. Claude suggests four plausible causes. Only her own scheme’s figures will settle which cause is at work. Professional next step?

Build Exercise: Plan It Twice, Then Find the Real Bottleneck (40 min)

CCAO-F Domain 4 · Task 4.3

What You Need to Know

Solution iteration improves a deliverable over several rounds, with Claude as the drafting partner. A solution here is a process, template, piece of guidance, scheme, service design, form or plan — never software.

Why the Loop Is Worth Running

Claude changes the economics: a second and third version cost minutes. Comparing three concrete options is a better way to design than arguing about one. It only works while the versions stay disposable.

One Named Criterion per Round

Every round needs a named criterion from the requirement. “Make it better” produces something different, not something better. Run each round against one criterion, check the result against that criterion alone, and keep the version that passes. (AI Fluency: Description-Discernment Loop.)

Knowing When to Stop

Stop when remaining changes are matters of preference rather than fitness. If a round shows the requirement itself was wrong, the next stop is the requirement, not another draft. State the stopping reason in one line.

Artifacts and Branches

Substantial content goes into an artifact. To compare two options, edit the message that produced the draft — that creates a different version of the chat with its own artifacts, while the original branch keeps the first draft. Keep the requirement visible while you compare.

Key Concept: When a stem shows a workable first draft, name one criterion from the requirement and check the revision against it.

Exam Traps

• Round after round of make-it-better with no named criterion
• Redrafting a deliverable whose requirement was wrong
• Letting the best-written version beat the one that fits

Practice Scenario

Laser-tag arena party coordinator is designing the running order for a new birthday-party package. Requirement fixed: ages 7–11, safety briefing before play, food in the party room, all inside 90 minutes. Claude’s first draft is workable. What should round two look like?

Build Exercise: Three Rounds, Three Criteria (35 min)

CCAO-F Domain 4 · Task 4.4

What You Need to Know

Integration here asks where in an existing process a person brings work to Claude. It never means a technical connection between systems.

Three Moves

Move	Meaning
Augment	Hand Claude one slow step; everything else keeps its shape. Cheap, reversible, easy to explain. The right choice far more often than the ambitious alternative.
Redesign	Change the process itself because the constraint that gave it its shape has gone. Pilot first (Anthropic’s own deployment guidance stages adoption this way).
Escalate	Pass work beyond the Associate’s role to a Claude Developer or Architect — systems talking to each other, data moving on its own, process running with nobody present.

The Human Gate

Find the step where the decision happens, and keep a person on it. Integrations fail when the checking step is removed. Checking has no visible output, so it looks like waste on a process diagram.

The Associate’s Boundary

Requests that need a technical implementation (automatically, without anyone having to, straight from the system, overnight) are escalated with a clear written description of what the process needs. Escalation is no failure — it is knowing the shape of your own job.

Where Recurring Work Lives

Recurring work belongs in a Project, not in one ever-lengthening conversation. Configuring the Project is Domain 5’s job; deciding the work belongs there is this objective’s.

Key Concept: Wiring Claude into a system or setting it up to run unattended is this domain’s signature distractor and is never the right answer.

Exam Traps

• Improvising an unattended pipeline that wires systems together
• Redesigning the whole process at once
• Optimising away the checking step

Practice Scenario

Local newspaper’s events page is compiled weekly by an editorial assistant from event details venues email in. A colleague proposes Claude should gather listings from venues’ websites overnight and publish the finished page without anyone touching it. The assistant still wants to save time on this week’s page, with a person reading every listing before it runs. How should the assistant respond?

Build Exercise: Map One Process You Own (45 min)

CCAO-F Domain 4 · Task 4.5

What You Need to Know

Say what is true, including the parts that are not impressive. The report serves two audiences: the enthusiast and the sceptic.

Stating Value in Four Parts

1. The task
2. The before
3. The after
4. How you know

“The monthly board pack takes two hours instead of five, measured across three months” is a claim. “Claude has transformed how we work” is not. Every figure appears because somebody measured it over a stated period. A measured saving belongs to its task — never projected onto work the trial did not touch.

The Untested Category

Limitations get stated plainly: what still needs a person and why, what went wrong during the pilot, and what hasn’t been tested. Silence about the untested category gets heard as a claim.

A Caught Error Is Evidence

A booking confirmation with the wrong date that the read-through caught is evidence the checking step works — exactly what the sceptic needed to hear. Hidden and surfaced later, every other claim gets re-read as curated.

The Enthusiast and the Sceptic

• Enthusiast (“can it handle all of it now?”) → name what stays human, not a flat no
• Sceptic → measured result, named human step, straight account of what went wrong and how it was caught

Under-claiming is not honesty. Overcorrection with so many caveats that a working result sounds like a failure hands the sceptic a conclusion the data does not support.

Key Concept: The only number that can support a keyed answer is one the stem itself says was measured.

Exam Traps

• Presenting enthusiasm as evidence
• Projecting a measured saving onto untested work
• Leaving the pilot’s caught error out of the report
• Burying a working result under caveats

Practice Scenario

Talking newspaper coordinator has trialled Claude for three months. Listener letters now reach a first draft in minutes rather than half an hour, measured across the trial, with every letter still read and approved by a volunteer. Edition scripts saw no saving. One draft gave the wrong date for the annual listeners’ tea and was caught at the read-through. Committee has asked for what the trial showed, with nothing projected or hypothetical and nothing left out. What should the report say?

Build Exercise: The Three-Part Account (30 min)