Analysis and hypothesis testing
Prepares repeat calculations, comparisons and scenarios for the specialist to interpret.
Deliver engagements
The saved work
Reviewed operational diagnostic
For 26 October 2026, 300 available hours at an estimated five hours per job model 60 jobs; this is not measured throughput. The confirmed-only comparison uses 60 jobs. The planning register still contains 80 jobs, including 20 provisional jobs dependent on client access. If none of those provisional jobs release during the week, the confirmed-only comparison remains 60; if all release, the potential workload is 80 against the same 60-job model. The planning register remains 80 in both scenarios. These conditional comparisons do not establish delivery readiness, spare capacity, staffing need or a commitment. Confirm actual access-release dates, validate job mix and measured effort, and compare scheduling and service scope before deciding.
The effort input is an estimate. Validate the period, job mix and available options before any commitment. This comparison does not establish staffing viability or guaranteed delivery.
A related local example. It shows the stated part of the work, not every step in this design. Read it in saved outputs
- Starts when
- Defined analytical question
- What it works from
- Evidence, models, assumptions and alternative explanations
- What it prepares
- Analysis workpaper, scenarios and unresolved tests
- What your team decides
- Specialist decides meaning and recommendation
What this helps your team do
This path has been checked in the fictional local build.
Responsibilities connected
See what each agent is responsible for preparing in this workflow. These are designs and local examples, not a set of independently running agents.
When it stops and asks
Missing/conflicting sources, unsupported calculations or unresolved judgement produce a partial result or explicit hold on the affected output.
Supported scope, source access and authority are agreed for each installation.