AI Prompts for Process Improvement

Process improvement is a method, not a mood. The method starts by seeing the process as it actually runs — every step, handoff, wait and rework loop — and measuring it, so that waste is identified from evidence rather than opinion. Then options are generated and evaluated against the goal, and the work is chartered with a baseline and a target so that success can be measured.

These prompts follow that sequence: a current-state map with the eight wastes and flow metrics, improvement options evaluated on effect, effort and risk, and a project charter in DMAIC form. They apply to manufacturing, warehouse, order-to-cash, procure-to-pay or any repeatable process.

Before you use these

Have these ready to replace the highlighted [variables]:

• A walk-through of the process as performed: steps, who, systems, time per step, waiting between steps, rework loops
• Volumes, cycle time, error rates, cost
• The problem or goal that triggered the improvement
• Constraints: systems, policy, headcount

The prompts

• 1. Map the current state and identify waste
• 2. Generate and evaluate improvement options
• 3. Write the DMAIC or kaizen project charter

1. Map the current state and identify waste

Act as a continuous improvement lead building a current-state map of [process].

Walk-through: [step, who performs it, system or tool, process time, wait time before it, batch size, error/rework rate, handoffs]
Volumes: [units per period] Demand rate: [takt if relevant]
Known pain points: [from the team]

1. Build the map as a table: step, actor, process time, wait time, value-add / necessary non-value-add / waste, information flow, rework loop (to which step, how often).
2. Compute: total lead time, total process time, flow efficiency (process ÷ lead time), number of handoffs, first-pass yield through the whole process.
3. Identify waste by category with specific evidence and estimated cost or time per period: overproduction, waiting, transport, over-processing, inventory/WIP, motion, defects/rework, unused skills. Do not list a waste without pointing to the step and the data.
4. Identify the largest queues and their causes (batching, capacity, approval, information missing).
5. Rank the improvement themes by time or cost recoverable, and note which require only method change versus system or policy change.
6. Data gaps: steps where time or error data is estimated rather than measured, and how to measure them.

Present the map table first, then findings. Keep interpretation separate from observation.

2. Generate and evaluate improvement options

You are generating improvement options for [process] to achieve [goal: metric, baseline, target].

Current-state findings: [wastes, queues, rework loops, flow metrics]
Constraints: [systems, headcount, policy, budget, timeline]
Levers available: [method changes, layout, batch size, standard work, visual management, error-proofing, system configuration, policy change, automation, role changes]

1. For each major waste or queue, generate 2–3 options using different levers. Each: mechanism (how it removes the waste), expected effect on the goal metric with the basis, effort (days/cost), risk (to quality, service, people), dependencies, and how it is sustained (standard, control, metric).
2. Evaluate on an effect/effort matrix; mark quick wins (high effect, low effort), projects (high/high), fill-ins and avoids.
3. Check interactions: options that conflict or that only work together.
4. Recommend a sequence: quick wins in the first 30 days, the project(s) to charter, and what to defer with the reason.
5. Estimate the cumulative effect on the goal metric if the recommended set is implemented, with a range.
6. Identify the option most likely to be resisted and why, and the mitigation.

Do not propose automation for a process that has not been simplified first. State the assumption behind every effect estimate.

3. Write the DMAIC or kaizen project charter

Act as an improvement program lead writing a project charter for [project name] on [process].

Problem: [what is happening, measured how, baseline value, cost or consequence]
Goal: [metric, target, by when]
Scope: [in and out]
Team: [sponsor, leader, members with roles, time commitment]
Approach: [DMAIC / kaizen event / PDCA] Timeline: [weeks]

Write the charter:
1. Problem statement: specific, measurable, no causes or solutions embedded.
2. Business case: cost of the problem, benefit of reaching the goal (financial and non-financial), link to site priorities.
3. Goal statement: metric, baseline, target, date; and any secondary metrics that must not deteriorate.
4. Scope: process start and end points, included and excluded areas, constraints.
5. Team and roles; sponsor's commitments.
6. Plan by phase (Define, Measure, Analyze, Improve, Control — or the kaizen event structure): deliverables, dates, tollgate questions.
7. Measurement plan: how the goal metric is captured, by whom, how often; data validity checks.
8. Risks and dependencies; stakeholders to engage.
9. Approval block.

Keep to one page. Flag if the goal is not measurable from current data and what must be put in place first.

Related prompts

• Root Cause Analysis
• Bottleneck Analysis
• SOP and Operating Routine Drafting
• Operational KPI Reviews
• Warehouse Operations Improvement
• Operational Cost Reduction

Logical next step

After this, most operations teams move on to Root Cause Analysis.

All Operations & Supply Chain prompts · Search the full library