AI for Business Analysts: From PRD to Workflow Design

Learn how to use AI to turn a product requirements document into a future-state workflow. I walk through how business analysts can challenge assumptions, identify missing steps, and validate solution design with their teams.

11/19/20262 min read

You created a product requirements document using AI. Now, how do you turn those documented requirements into a proposed solution that your team can understand and evaluate?

This is where business analysts contribute to solution design. In my tutorial, I use Claude to draft a future-state workflow for a hypothetical grocery delivery application. Then I challenge the output to identify missing steps, unsupported assumptions, and questions requiring stakeholder decisions.

The important part is knowing what to question before accepting the workflow.

Start With the Requirements You Already Have

For this example, the business problem is limited visibility into grocery deliveries. The objective is to improve the customer experience and reduce avoidable tracking inquiries.

One functional requirement says customers should be able to view their current delivery status.

That requirement gives us a starting point, but it does not explain every step. How does the customer select an order? Where does the status come from? What happens when information is unavailable?

Using AI for business analysis helps us explore these questions while keeping the documented requirements as our source.

Ask AI to Draft a Future-State Workflow

I asked Claude to identify the actors, actions, decisions, handoffs, and alternate paths using the existing PRD.

I also asked it to separate supported steps from suggestions and unresolved questions. Each proposed step should connect to a requirement or business objective.

Why does that matter? Because a workflow can look complete while including business rules nobody approved.

AI can help organize the analysis. The business analyst still needs to check whether the proposed behavior is justified.

Challenge What AI Generates

When reviewing Claude’s workflow, I noticed several areas that needed closer attention.

The process sequence: Claude mixed background status updates with the customer’s tracking journey. We needed to clarify where the customer workflow begins and how supporting system activities connect.

Order selection and authorization: The workflow said the application retrieves the status “for that order.” But which order? What if the customer has several? How does the application verify that the customer can access it?

Outdated information: An available status is not necessarily current. If the application shows “being prepared” after the driver has collected the groceries, what should happen?

These questions reveal gaps. They do not give AI permission to invent the answers.

Separate Business Goals From Workflow Decisions

Claude also asked whether the displayed status answered the customer’s question.

But how would the application know that?

Reducing support inquiries is a business objective we need to measure. Displaying a status does not automatically mean the customer has everything they need.

I also questioned whether support agents handling inquiries belonged in the core tracking workflow. Keeping those supporting activities separate helps prevent unplanned scope expansion.

Validate Before Moving Forward

An AI-generated workflow remains proposed until the responsible people review it.

Product and business stakeholders confirm policies and expected behavior. Operations and support review practical scenarios. Engineering assesses feasibility and dependencies. Design partners evaluate the experience when appropriate.

The business analyst documents decisions, unresolved questions, and requirement updates. Confirmed behavior can then inform user stories, acceptance criteria, and backlog refinement.

My approach to AI-assisted business analysis is straightforward: provide context, challenge assumptions, and validate decisions. That is how AI can support better solution design and Agile product delivery.

Watch the full tutorial for the step-by-step example.

Nan Ross

Agile Product Delivery & AI Adoption Expert. Helping leaders and teams turn ideas into working products with clarity, not chaos.

© 2026 Corporate Cosmo, LLC DBA NanRoss.com. All rights reserved.

My Books

Connect

Work with Nan

Projects