ChatGPT & Me Episode 3: Analyzing for logic and clarity

In Episode 2, I learned ChatGPT can generate flowchart code. The code can be uploaded to the Python GPT, which converts it to a flowchart diagram.

For this episode, I converted to ChatGPT PLUS and tested the cycle of flowchart code —>flowchart diagram and added an analysis step. Actions taken included:

• asked ChatGPT to create Graphviz flowchart code from a numbered list of steps representing a quote-to-sales-order process

• uploaded the code to Python which generated the flowchart diagram

• provided ChatGPT with the flowchart diagram and asked it to analyze the logic

ChatGPT produced a list of questions even a new hire would want answered. The process of answering the questions could reveal new best practices, improving the workflow and employee morale. If a team already has workflow diagrams, and has access to a company-approved AI tool, it can request a review of the workflow’s logic and expect to receive valid, actionable feedback.

Workflow diagram provided to ChatGPT

Diagram was created by the Python GPT from code I requested from ChatGPT. It’s not polished, but it’s a good start.

ChatGPT’s Analysis of the workflow

I have reviewed the workflow diagram, and here are some questions regarding its logic and clarity:

1. Missing Information Handling:

   • If missing information is identified, is there a limit on how many times the request can be sent back before it is rejected or escalated?

2. Manufacturer Follow-Ups:

   • How long do you wait before following up with the manufacturer if the quote request is not confirmed?

   • What happens if the manufacturer never confirms the quote request?

3. Sales Rep Review:

   • What criteria define whether a quote is "Good" or needs revision?

   • If a quote is revised multiple times, is there a limit on how many revisions can be made before it is reconsidered?

4. Contractor Approval Process:

   • If the contractor does not accept the quote, what happens next? Is the process restarted or does it move to a different resolution step?

   • How long do you wait before following up with the contractor?

5. Sales Order Creation:

   • What happens if a sales order number is not created? The diagram mentions "Troubleshoot," but does not specify what actions are taken.

   • Who is responsible for troubleshooting, and is there an escalation path?

6. Manufacturer Delivery Confirmation:

   • If the manufacturer does not confirm the delivery date, is there a threshold for how many follow-ups occur before alternative actions are considered?

7. Order Delivery Issues:

   • What constitutes an "address issue" if the incorrect order is delivered? Does this mean returning the order, requesting a new shipment, or some other corrective action?

   • Who is responsible for resolving delivery issues—manufacturer, contractor, or internal teams?

8. End Conditions:

   • Is there any tracking mechanism for completed orders to ensure successful delivery and invoicing?

   • What happens if an order is delayed significantly—does it go back into the workflow at an earlier stage, or is there a separate resolution process?

  Operational excellence begins with clarity on workflows and procedures.

This blog was co-written by a human, ChatGPT, and Python GPT and edited by a human!

© 2025 Lori K. Barbeau