The 5 As of AI-enhanced Agile

The 5 As of AI-enhanced Agile

Use these basic prompts to apply the 5 A’s of AI-Enhanced Agile. Test them, tweak the wording, and try different AI tools—each may give you new insights. Experiment, refine, and see what works best for your workflows.

Apply

Assess

Align

Adjudicate

Advance

Assess

Back

Try this prompt!

Here's an example

"You are a process improvement analyst. Review this workflow: [insert workflow]. Identify where AI could save the most time, reduce costs, or improve accuracy. Provide at least two high-value opportunities, explain the potential impact, and recommend which should be prioritized first."

Apply

Back

Try this prompt!

Here's an example

"You are a pilot design consultant. I want to test AI in a limited workflow: [insert workflow]. Create a pilot plan that defines scope, success metrics (e.g., time savings, error reduction, quality improvements), and evaluation methods. Recommend how we should decide whether to scale the solution across other workflows."

Adjudicate

Back

Try this prompt!

Here's an example

"You are a decision-support advisor. AI produced the following output for prioritizing tasks: [insert AI-generated list]. Review these recommendations against practical considerations such as strategic goals, resource availability, and operational constraints. Highlight where human oversight should override AI suggestions and explain why."

Align

Back

Try this prompt!

Here's an example

"You are an AI governance consultant. AI suggested: [insert recommendation]. Evaluate the recommendation for fairness, compliance, and quality risks. Provide a transparent explanation of how the AI might have generated this output, and suggest governance guidelines that ensure decisions remain accountable and auditable."

Advance

Back

Try this p rompt!

Here's an example

"You are a capability-building consultant. The team has limited AI experience. Create a short upskilling plan with three parts: (1) a simple task to practice using AI, (2) a reflection activity to compare AI vs. human-generated outputs, and (3) guidance on how to spot when AI recommendations may be inaccurate or biased."

Example:

“You are a capability-building consultant. The team uses an AI-powered knowledge base for answering customer questions. Create a 3-part training activity:

1. A simple exercise where agents use the tool to answer 5 sample customer queries.
2. A reflection task comparing AI-generated answers vs. human-written answers for accuracy and tone.
3. Guidance on spotting outdated or biased AI responses and how to escalate them for correction.”

Example:

“You are an AI governance consultant. The AI recommends shortening average call duration by removing certain verification questions. Evaluate this for (a) compliance risks, (b) fairness to customers, and (c) potential impact on service quality. Explain in plain language how the AI might have generated this recommendation, and propose 2 governance steps to ensure transparency and accountability.”

Example:

“You are a decision-support advisor. The AI predicts a 30% increase in support tickets during monthly billing cycles. Compare this with human/organizational considerations such as staff availability, skill sets, and client priorities. Suggest 2–3 ways humans should validate or adjust the AI recommendation before taking action.”

Example:

“You are a pilot design consultant. We want to test AI in automated email categorization for one client account. Create a pilot plan that defines:

1. The scope (which emails are included/excluded),
2. At least 2 success metrics (e.g., % faster routing, % fewer misclassified emails),
3. How results will be measured after 30 days, and
4. A clear decision rule for whether to expand the solution to other client accounts.”

Example:

"You are a process improvement analyst. Review this workflow: customer service agents handling incoming calls, logging tickets, escalating issues, and preparing daily reports. Identify 3 tasks that are repetitive or time-consuming and explain how AI could (a) reduce manual effort, (b) improve accuracy, or (c) speed up completion. Rank these opportunities by potential impact and provide a recommended first step."

Here's a pro tip!

For any of these prompts, the more context you provide—such as volumes, timelines, roles, or success criteria—the more accurate and tailored the AI’s output will be. Don’t hesitate to test, adjust wording, or add details until the results fit your needs.