LibraryComparisons

BPMN with ChatGPT

A general chat assistant such as ChatGPT can explain BPMN, describe a process in prose and, if you ask, write BPMN 2.0 XML as text. That is useful, and this page says where. patchley is built for one job: it listens to the interview, drafts the diagram with lanes for the roles, keeps every version, writes the analyses per step and fills the deck. The comparison is between a general tool and a specialised one.

Side by side

A general tool and a specialised one

The left column describes a general chat assistant as a category: a text conversation with an AI model. The right column states what patchley does.

General chat assistant (ChatGPT and similar)patchley
How the process gets inWhat you type or paste into the chat: a description, a list of steps, an existing text.From the conversation: live audio in a workshop, typed text, a PDF or DOCX up to 50 MB with optional focus, audio files (MP3, WAV), a BPMN file, or patchley joins a Teams, Zoom, Webex or Google Meet call as a visible participant.
What you getText: an explanation of the notation, a process described in prose or as a list, and on request BPMN 2.0 XML written out as text that you copy into a modelling tool.A BPMN 2.0 diagram on a canvas, with pools and lanes for the roles, systems, documents and control steps marked; a process landscape from several diagrams.
Editing and changingYou ask again and get a new text. Keeping track of versions, and of what changed between them, is up to you.On the canvas or by describing the change. The analyst proposes, every change is shown before it lands, your own edits are never overwritten, every version stays.
Beyond the diagramFurther text on request: a summary, a draft SOP, a list of questions to ask.Analyses per step (SOP, pain points, compliance, quantitative figures, your own template) and PowerPoint decks filled from the diagram and the analyses.
Exchange formatsText you copy out of the chat, including BPMN XML as text. Whether a file download is offered depends on the provider and plan.BPMN 2.0 XML (opens in Camunda Modeler, Signavio, ADONIS), SVG and PNG for the diagram, XLSX for the analyses, PowerPoint for the deck.
HostingDepends on the provider and the plan; see the provider's site and terms.App hosted in Germany, AI models from a European provider, hosted in the EU.

When to use which

Where a chat assistant is genuinely useful, and where the job changes

The dividing line is not the quality of the AI. It is whether the process already exists as text in your head, or still has to be captured from the people who run it and agreed with them.

Chat assistant fits

Learning and explaining the notation

"What is the difference between an exclusive and an inclusive gateway?" A chat assistant answers this well and patiently. So does the BPMN guide, without a login.

Chat assistant fits

A first sketch of a process you already know

If you can type the steps from memory, a chat assistant turns them into a tidy list or into BPMN XML as text that a modelling tool can open.

patchley fits

Capturing a process from the people who run it

Four people in a workshop, each knowing one part, interrupting each other. The diagram has to grow while they talk, with lanes for their roles, and be corrected on the spot.

patchley fits

A diagram to sign off, analyse and present

Versions people can compare, control steps marked, analyses per step, and a deck for the steering committee, all from the same model.

In one sentence

A chat assistant is a good place to understand BPMN and to sketch what you already know; a process that has to be captured from several people, agreed, versioned and presented is a different job, and patchley is built for that one.

What is different in practice

Text in, text out, versus a model that is worked on

With a chat assistant, you are the interface: you type what you know, read what comes back and carry it into a modelling tool yourself. The roles, systems and documents in the result are the ones you mentioned; the chat does not know your organisation unless you tell it. That is fine for a sketch and becomes work when the process has forty steps and three people disagree about step twelve.

patchley works on a model, not on text. The conversation is the input; the diagram, the landscape, the analyses per step and the deck are views of the same stored process. When a participant says "no, the underwriter checks that before registration", the change is proposed on the canvas, shown, and only then taken over. How that generation works, and what you still decide, is described on BPMN with AI.

Chat assistant text

  • You describe, it answers
  • BPMN XML as text, copied out
  • Versions: your own files

patchley model

  • The workshop speaks, the diagram is drafted
  • Lanes, systems, documents, control steps on the canvas
  • Every version kept; analyses and deck from the same model
See this on one of your own processes.Book a demo

Book a 30-minute demo

We model it live with you, in your language. Pick a time and the Microsoft Teams invitation is in your calendar straight away.

Download request

Request a document

Enter your email address and we send you the document by email.

Requested document

Fields marked * are required.

See patchley in 19 clicks