AI Workflow Candidate Scorecard

A practical screen for deciding whether a workflow deserves deeper review.

How to use this resource

Choose one workflow. Score each dimension from 0 (not established) to 4 (well established with evidence), and record the evidence behind every score.

Important: A score is a planning aid, not legal, security, privacy, procurement, technical or financial approval.

Audience

Operations, technology, transformation, product, finance and risk leaders

When to use it

A specific workflow is being considered for AI-enabled improvement.

Objective

Identify evidence gaps and the appropriate next decision.

Business impact — material effect on cost, cycle time, quality, capacity, revenue or risk.
Volume and repetition — occurs often enough for improvement to matter.
Workflow stability — main steps, decisions and exceptions are understood.
Baseline evidence — current performance can be measured.
Data access — authorized, reliable data is available.
Integration feasibility — required systems can be accessed through controlled interfaces.
Evaluation method — representative test cases and failure criteria can be defined.
Human authority — approvals and escalation are explicit.
Risk and reversibility — errors can be detected, contained and reversed.
User adoption — frontline users will participate.
Operating ownership — monitoring, incidents and change have an owner.
Decision timing — a real sponsor, deadline and budget path exist.

0–15: Evidence first — clarify the workflow, baseline, ownership and data before selecting technology.
16–27: Readiness work — material dependencies remain.
28–39: Pilot candidate — confirm scope, acceptance and controls before testing.
40–48: Strong candidate subject to technical, security, privacy, legal and operating review.

Start with the problem

You do not need a specification. Tell us what you are trying to improve.