How to use ChatGPT for Efficient Business Process Modeling: A Guide

In my previous post, I discussed the potential of generating diagrams with ChatGPT. Today, I’ll dive deeper into how you can leverage ChatGPT for business process modeling effectively, ensuring you save time while maintaining accuracy in your diagrams. 

What you need for Effective Modelling

To create a process flow or any diagram using ChatGPT, here is what’s essential:

1. You need to have a solid process description.
2. You need to know that there is a diagramming language like Mermaid.
3. You need to have a Mermaid interpreter.
4. And of course, you need a modeling tool. 

Generate a solid process description

Let’s go back to the original lost cats and dogs platform. First, we need to specify a process for it. For this example, let’s ask ChatGPT to generate a business process for the platform, from registering a lost pet to the situation where the pet is found by another user, and the pet owner, pet, and finder are brought together, and everyone lives happily ever after. (In real life, I would suggest using your BA skills to describe and create the best-fitting process description and optimize it from there with ChatGPT.)

In response to the question above, you’ll get something like this:

“Let’s outline a business process from the point a lost pet is registered on your platform until the happy moment the pet is found and reunited with its owner. Here’s a simplified process:

1. Pet Owner Registers Lost Pet: The pet owner accesses the platform and submits details about their lost pet (photo, name, type, breed, sex, description, last seen location, etc.), along with their contact information.
2. Pet Registration Confirmation: The platform confirms the registration and displays the lost pet details on a public-facing page, providing the owner with a shareable link.
3. Notification to Community: The platform sends alerts to nearby users and posts on connected social media services to maximize the search area.
4. Finder Reports Found Pet: A user (finder) who identifies the pet matches it with the lost pets on the platform and reports finding the pet through the platform.
5. Verification of Pet Match: The platform verifies the match by comparing the details provided by the finder with the lost pet’s registered details.
6. Contact Exchange: Once verified, the platform facilitates the exchange of contact information between the pet owner and the finder.
7. Reunion Arranged: The pet owner and the finder arrange a safe meeting to reunite the pet with its owner.
8. Success Confirmation: After the pet is reunited, both the finder and the pet owner confirm the reunion through the platform.
9. Feedback and Review: Both parties are invited to provide feedback on their experience, which helps improve the platform.
10. Close Case: The case is closed, and the pet’s status is updated on the platform to prevent further alerts. “

What you need to know about Mermaid

In short, Mermaid is a syntax for an interpreter that allows creating diagrams from certain syntax. More can be found here. All you need to do is ask ChatGPT to transform your process description into Mermaid and you will get something like this:

 graph TD

A[Pet Owner Registers Lost Pet] –> B[Pet Registration Confirmation]
B –> C[Notification to Community]
C –> D[Finder Reports Found Pet]
D –> E[Verification of Pet Match]
E –> F[Contact Exchange]
F –> G[Reunion Arranged]
G –> H[Success Confirmation]
H –> I[Feedback and Review]
I –> J[Close Case]

The next thing you need is a Mermaid interpreter.

Therefore, you can use: Mermaid.live

All you have to do is copy and paste this code into Mermaid.live, and there you go.

so now the fun part starts; instead of just drawing and remodeling, we will challenge ChatGPT to improve the flow.

For example, we need to add a finder registration, right? And what if the found pet isn’t the missing pet? We might need to set up a process for registering a pet found not being the owner.

 We get something like this:

 graph TD

A[Pet Owner Registers Lost Pet] –> B[Pet Registration Confirmation]
B –> C[Notification to Community]
Z[Finder Registers as Finder] –> D[Finder Reports Found Pet]
D –> E[Verification of Pet Match]
E –> F{Pet Matched?}
F –>|Yes| G[Contact Exchange]
G –> H[Reunion Arranged]
H –> I[Success Confirmation]
I –> J[Feedback and Review]
J –> K[Close Case]
F –>|No| L[Register Found Pet Not Listed]
L –> M[Publishes Found Pet Details]
M –> N[Continues Search for Owner]
 

Copy pasting it in to mermaid.live get’ll get something like this.

So by this point you might thing just a flow chart isn’t suitable for this kind of process, so you will ask can you help me generate a sequence diagram  for this process for the following actors:

Pet Owner, Finder, Platform.

You ‘ll get something like this:

Or Experiment with one of the other diagrams supported by Mermaid

At a certain moment, you won’t be able to challenge the model anymore. By that time, it is time to wrap up.

Take your favorite modelling tool.

You will take your favorite modeling tool and will just copy-paste the model. If the tool supports Mermaid.js like Draw.io, then you just import it and tidy it up as you have to redraw it, but you only have to do it once.

This is how you can use ChatGPT for creating diagrams. Hope you will be able to apply this in your everyday business analysis work.