Enquiry triage
Route incoming questions and prepare a reviewed response.
Find the workflow worth improving first. Compare opportunities, clarify blockers and build a practical 30/60/90-day plan.
No signup · Runs in your browser · No AI calls
Choose the functions and friction you want to explore. These highlight templates; they do not change scores.
Add up to ten workflows. Templates start with unknown answers so you can assess your actual situation.
Route incoming questions and prepare a reviewed response.
Extract invoice details and flag exceptions before approval.
Find answers with references in approved internal sources.
Build a first draft from approved examples and requirements.
Assemble recurring reports from agreed sources.
Coordinate checklists, access requests and useful guidance.
Turn approved notes into reviewed actions and owners.
Sort incoming documents into agreed categories for review.
Use evidence where you have it. Leave unknowns visible. A missing answer is different from a negative answer.
Choose a workflow in step 1 to begin.
Compare the work. Then choose the next step.
Rubric 1.0 · 0 candidates. Input basis: No function or pain point selected.
Planning scores from your answers. Verify the evidence before making a pilot decision.
Add a workflow or try a fictional preset to see your map and roadmap.
Illustrative rubric, not a validated benchmark or professional audit. Answers are self-reported, not verified. All answers have equal weight within each dimension. No monetary savings are estimated. The 20-case baseline and 30/60/90-day plan are starting assumptions, not delivery commitments.
Illustrative rubric, not a validated benchmark or professional audit. Answers are self-reported, not verified. All answers have equal weight within each dimension. No monetary savings are estimated. The 20-case baseline and 30/60/90-day plan are starting assumptions, not delivery commitments.
Each anchored answer contributes 0, 1 or 2 points in the displayed order. Value averages frequency, time burden and reach. Effort averages systems, exceptions and integration. Readiness averages source quality, source permission, consistency and ownership. Risk averages sensitivity, external actions and error consequences. Each dimension is points ÷ maximum points × 100, rounded to an integer. Existing software and task type select the approach, not the score.
Unknown contributes 0 to value/readiness and 2 to effort/risk; it always remains unknown in the input basis. Completeness is answered fields out of 15, including task type. Not applicable is available only for systems, integration and external action: it explicitly means no such effort/action and counts as an answered zero, not missing evidence.
Pilot eligibility requires every answer to be known, source access and ownership both at 2, source quality and consistency at least 1, exceptions below 2, and all risk answers below 2. Eligible candidates rank first; within each group the score is 45% value + 25% (100 − effort) + 30% readiness. Ties use the candidate's stable ID in ascending order. Risk gates cannot be offset by value.
Approach order: unreliable sources, inconsistent process or frequent exceptions → improve the process; remaining blockers or unknowns → investigate; otherwise an available software feature → configure it; rules-based work → conventional automation; language/mixed work → AI assistance with review. Changing the scope or controls requires a fresh assessment.
No uploaded data is sent to a model, chat or contact service. Sharing uses an encoded URL fragment, which is not sent with the page request. Recipients and software handling the URL can still read it. JSON files contain the complete editable model. The PNG contains the top three opportunities and their answer basis. Print includes all candidates.