SAP systems have data related to business processes while the Microsoft Graph has data related to collaboration such as emails, chats, schedules etc. To use both these forms of data together with AI would transform the business processes of many organizations and make them more efficient.
Microsoft offers powerful AI and low code development tools that are easily connected to SAP systems to develop your own intelligent copilots and bots to enhance your business processes and workflows.
To demonstrate how seamless and easy it is to utilize AI using Microsoft AI tools, we will explore a scenario that demonstrates how SAP data along with Copilot Studio, Azure OpenAI, Microsoft Teams, and the Microsoft Power Platform can be leveraged to streamline a business process. There are two scenarios 1 being a basic one and 2 with enahnacements to the first one inlcuding Adaptive Cards and Principal Propagation.
Be sure to read this blog on the Microsoft Tech Community for a more detailed description of this scenario.
A customer contacts a salesperson with a complaint about a delayed order.
The salesperson uses a chatbot deployed on Microsoft Teams to investigate the customer's orders by leveraging AI to gain summarized insights from Outlook Emails and Teams’ messages, combined with direct data from the SAP system.
After identifying the issue, the salesperson uses AI to generate a report summarizing the conversation and the cause of the delay, which is then sent to the customer via email through the chatbot interaction on Teams.
Be sure to read this blog on the Microsoft Tech Community for a more detailed description of this scenario.
The salesperson asks the Copilot to help her look through all the materials in the SAP system to find the best replacement.
The salesperson gets a suitable replacement suggestion from Copilot and tries to access material stock information for that material. However, the salesperson is not authorized to do so as she does not have the right authorization to do so in the SAP system. The salesperson messages her colleague who does have the authorization to check material stock information for her. The colleague checks the stock information and informs the salesperson that the material is in stock.
The salesperson now decides to update the sales order with the new material and remove the old material from the sales order.
There are two SAP connectors on the Power Platform that can be made use of to recreate this scenario:
1)The RFC connector:
2)The new SAP OData connector with SSO using Microsoft Entra ID:
This scenario showcases the integration of various Microsoft technologies, including:
- Copilot Studio
- Azure OpenAI
- Microsoft Teams
- Microsoft Power Platform
This repository contains Power Automate flows and a detailed PDF with steps on how to build out this scenario using these flows in Copilot Studio.
Follow the detailed steps provided in the PDF to set up the scenario in Copilot Studio and customize it according to your business needs. The PDF is named "Steps to recreate the scenario.pdf" for scenario 1 on this repo, make sure to download to PDF to access the links. For scenario 2, the instructions are in the PDF "Steps to recrate scenario -2 . pdf" and the flows and JSONs (for the adaptive card) have a commit description of "Scenario 2 flows".
The steps to recreate the entire scenario using the SAP OData connector are in the PDF named "Steps to recreate the Scenario - OData"
- Go to my flows in Power Automate and Click on import, under import click on "Import Package".
- Upload the Zip folder
- Once the upload is complete, create the own respective connections and add them to the flows by clicking on "Select during import" and once that is done click on 'Import' at the bottom.
- Since the SAP server information has not been entered in the flow, the flow wont be ready to use and you will see the image, click on 'save as new flow' and edit the information in the flow editor.
Note : Once you import the Power Automate flows, you will be prompted on the Power Platform to make an SAP ERP connection.Therefore, when you attempt to recreate this flow, complete the prerequisite steps highlighted in the PDF before you try to import the flow.(The PDF is called 'Steps to recreate scenario.pdf' on this repo.) Each flow is named after the flow mentioned in the PDF and are the .zip folders on this repo.
If you face issues while replicating the scenarios or have suggestions on new features/flows you would like to see, you could use GitHub Issues to report errors or request new features.