1
of 6— Gather your inputs before going to AI
What you'll accomplish
By the end of this guide, you'll be using Claude to produce polished process improvement recommendation reports — the kind that get stakeholder buy-in — in a fraction of the time. You'll learn how to structure your analysis input so AI produces professional-quality output you can send directly to a sponsor.
What you'll need
- Claude (free tier at claude.ai — or Pro for very long processes)
- Your process analysis notes: current-state description, pain points, potential solutions
- Time needed: 15 minutes to learn; 30–60 minutes per recommendation document vs. 4–6 hours previously
- Cost: Free (Claude free tier handles most process docs)
How-To Guide: Write Process Improvement Recommendations with AI
Step 1: Gather your inputs before going to AI
Before prompting, spend 10 minutes collecting:
- Current process description (step-by-step, even if messy)
- Top 3–5 pain points (time wasted, error rates, bottlenecks, handoff delays)
- Potential solutions (even rough ideas: automate step 3, eliminate approval for low-value items, consolidate two systems)
- Constraints (budget limit, technology stack, timeline, regulatory requirements)
The quality of your AI output directly correlates to the quality of this input.