Updated: November 2026

AI for ERP discovery workshops and requirements

AI notetakers already record ERP discovery workshops. The hours go into the next step: turning many calls, email threads and documents into requirements per business process that the client signs, with the open questions and decisions tracked. This guide shows which AI tools help with which part of discovery, where each is stronger and where it stops.

Record
As-is process
Require­ments
Open questions
Decision log
Meeting notetakers
One meeting at a time
Covers Record.
SAP Signavio, Opkey
From current system data
Covers As-is process.
NetSuite SuiteExpress
NetSuite only, in NetSuite's services
Covers Requirements.
superglue
Reads your notetaker's transcripts, threads and documents
Covers Requirements to Decision log.
Which part of ERP discovery each tool covers, November 2026.
Covers

What discovery has to produce

  • Requirements per business process: each with its source and the person who stated it.
  • Open questions and conflicts: for the next workshop, before they become design mistakes.
  • A decision log: who decided what, and where, so design and testing can reuse it.
  • The as-is picture: how the client works today in its current systems.

Meeting notetakers

Copilot in Teams and other notetakers record each workshop, write a recap and answer questions about the meeting. Teams recaps need a Microsoft 365 Copilot license or Teams Premium.

Stronger at: a reliable record of every call, with no extra work in the meeting.

Where they stop: each summary covers one meeting. Requirements that span several workshops, emails and documents still have to be pulled together by hand.

Process mining: SAP Signavio and Opkey

The SAP Signavio Process Insights discovery edition analyzes how an SAP customer uses its current system, as a one-time assessment that SAP customers can request for free. Opkey captures existing workflows and maps them to standard ERP processes.

Stronger at: the as-is picture from system data, before the first workshop.

Where it stops: it shows what the client does today, not what it needs from the new ERP.

Vendor AI: NetSuite SuiteExpress and SAP Joule for Consultants

NetSuite SuiteExpress builds a customer knowledge base from meeting notes, documents, spreadsheets and recordings, and its AI captures the requirements inside NetSuite's own implementation services. SAP Joule for Consultants answers SAP product questions for the consultants running the workshops.

Stronger at: projects that stay inside one vendor's product and services.

Where they stop: SuiteExpress covers NetSuite in NetSuite's services, and is live first for software companies in North America. Joule answers questions about SAP, not about the client.

superglue

superglue is an agentic implementation platform for partners and consultants. Its agent reads the workshop transcripts from your notetaker together with the email, Slack or Teams threads and the documents of the project. It writes the requirements per business process with the source and the person who stated each one, lists open questions and conflicting statements for the next workshop, and keeps a decision log that design, testing and your colleagues reuse.

  • Works with the notetaker you already use
  • Works for SAP S/4HANA, NetSuite, Sage Intacct, Business Central, Acumatica and other ERPs
  • The client gets a requirements document per business process and a decision log with sources

See how it works for SAP S/4HANA, NetSuite and Sage Intacct.

Where others are stronger: notetakers for recording the calls, process mining for as-is data from the current system, and Joule for SAP product questions.

Side-by-side (November 2026)

ToolPart of discoveryStarts fromProduces
Meeting notetakersRecording the workshopsOne meetingA recap per meeting
SAP Signavio, OpkeyAs-is analysisData from the current systemAs-is processes compared with standard
NetSuite SuiteExpressRequirements inside NetSuite's servicesNotes, documents, recordingsA customer knowledge base and requirements
SAP Joule for ConsultantsProduct questionsConsultant questionsAnswers with source documents
superglueRequirements and decisions across the projectTranscripts, threads and documentsRequirements per process, open questions, decision log

Try it on one client project

Give the agent the transcripts and threads from one client's discovery and review the requirements it writes: start your free trial. The next step is covered in how to speed up ERP fit-gap analysis with AI.

Common questions

How can AI help in an SAP discovery workshop?

Before the workshop, process mining such as SAP Signavio shows how the client uses its current system. During the workshop, a notetaker records what was said. After it, AI turns the transcripts, emails and documents into requirements per business process, lists open questions and conflicting statements for the next workshop, and keeps a decision log. SAP Joule for Consultants answers SAP product questions along the way.

Can AI write ERP requirements from meeting notes?

Yes, if it has all the sources. One meeting summary is not enough: requirements are spread across several workshops, email threads, chat and documents. AI that reads all of them can write the requirements per process, with the source and the person who stated each one, and flag where two people said different things.

Is a meeting summary from Teams Copilot enough for discovery?

A summary covers one meeting. Discovery needs the requirements across all meetings and threads, the open questions, the conflicts and the decisions, in a form that design and testing can reuse. Use the summary as one input, not as the requirements document.

What does the client sign at the end of discovery?

Usually a requirements document per business process, with the scope and the priorities agreed. The decision log records who decided what and where, so later phases can trace each design choice back to a requirement.

What gets lost between pre-sales and delivery?

Promises made in sales calls, edge cases mentioned once in an email, and decisions taken in chat. When the requirements are written from every source with a link to each one, the delivery team sees what pre-sales heard.