Model Driven PowerApp: 360 Degree View Dashboard to Monitor Data from External Systems (Azure AD and Azure Cosmos DB)

Problem: The business dealership information is being captured on multiple external systems. There is a need for business to capture 360° data view dashboard to keep track of information of dealers, sales data, and many other dealership information at one place. And this dashboard is for business users on the organization for monitoring.

Let us take an example of dealership architecture, where dealer identities/domain information are stored on external Azure AD, and basic details of dealership is stored on Azure Cosmos database.

Note: The dealership use case is an example for us to explore the possibilities. Similarly, this can be replaced with any other data model.

Design & Solution Considerations

The following elements/components are considered for building this solution.

• Azure Cosmos Database, which holds dealership’s basic information 
• Azure Active Directory, which holds the domain/identity information of dealership users [This is external/separate domain, holding only dealership users] 
• Power Automate, to integrate and push the data to CRM system. 
• Microsoft Common Data Service, which acts as intermediate storage and containing subsets of information from two other systems. [Azure Cosmos DB and Azure AD] 
• Power Apps – model driven app, which has necessary dashboard for business users. 

The following shows high level design of 360 degree architecture, integrating data from multiple systems.

High level design for Dashboard providing 360 degree of dealership data

The following illustrates the design.

• The necessary data model or entities are created on Microsoft CDS to capture the data from multiple system. For now in this usecase, let us focus on one entity, I.e., dealers.  
• The flow configured on Power Automate runs as a scheduled job, to synchronize the data into CDS. Pulls minimal information from Azure AD and Azure Cosmos database, and synchronize the data subset into Microsoft CDS.  
• Power Apps, which has views, forms and dashboards pulls and shows the information from entities, which is configured in the underlying CDS. 

Now let us get deeper into the solution, to see how these are configured.

Provision Private Channels on Microsoft Teams with Tabs

Let us understand how to auto-provision a private channel on Microsoft Teams, along with necessary tabs added on the private channel in an automated way.

Article: Private channel provisioning along with necessary tabs could automated with the help of Power Automate flow and Microsoft Graph API. Microsoft Graph API provides us the beta/major version of the graph endpoints to achieve the provisioning solution. Power Automate flow is configured with necessary HTTP requests for provisioning and automate the process.

The solution requires two major steps.

• Creating Private Channel
• Adding Tabs

Create Private Channel on Microsoft Teams 

At the time of writing this article, private channel creation endpoint is not available with major version (v1.0). Major version of graph endpoint, provides option to create a channel, but not private channels.

POST https://graph.microsoft.com/v1.0/teams/{id}/channels
Content-type: application/json
{ "displayName": "Architecture Discussion", "description": "This channel is where we debate all future architecture plans" } 

The beta version of graph endpoint, provides option to create a private channel. The endpoint requires a payload to define the nature of channel and to add members with their roles. The following snapshot from Power Automate shows the way to create a private channel, with a member and their role.

Create a private channel

Configure Azure AD end-user Authentication and Office 365 data access on Azure Health BOT Service via Graph API

Microsoft Health BOT Service is hosted on Azure, and is easily configurable solution that might meet multiple health-related BOT scenarios.

The scenario considered here in this article, helps us to understand how to enable end-user authentication on BOT Service to read Office 365 relevant data. For this POC, Azure AD authentication is considered. Azure AD authentication is best, only when the BOT service provider/organization manages their users entirely on Azure AD. If not the identity and authentication to be changed accordingly.

So in this article, let us understand how to enable end-user authentication on Azure Health BOT service using Azure AD, and read user profile information using Microsoft Graph API.

Enable Azure AD Authentication Flow using Azure AD App configuration

Create an application on Azure AD, and configure the following.

• Set the required API permissions. In my case, only User.Read under Microsoft Graph delegated type.
• Set oauth2AllowImplicitFlow to true, under Manifest section or platform configuration section.
• Under platform configurations, configure the redirect URI (https://bot-api-eu.healthbot.microsoft.com/bot/redirect/oauth2) for health BOT service authentication. 
• Generate the secret key from certificates & secrets. (Dem81hW@h:Ia07Os/kQEzk-Cr48_SuLm)

Copy the client ID and client secret from this app, to be used in the next sections on Azure Health BOT service configurations.

Configure Azure AD Authentication on Health BOT service

From the health BOT service (https://eu.healthbot.microsoft.com/account/<account-id>), navigate to integration -> authentication.

Authorization URL: https://login.microsoftonline.com/<tenantname>.onmicrosoft.com/oauth2/v2.0/authorize
Access Token URL:  https://login.microsoftonline.com/<tenantname>.onmicrosoft.com/oauth2/v2.0/token

Add new authentication provider with the configurations below.

Azure AAD Authentication configured from the configurations done on Azure AD App

Azure Web App – Integrate Microsoft Teams Channel and show Conversation highlights with MS Graph API endpoints

This article helps you out integrating the Microsoft Teams channel data into custom web applications. Assume an use case, web application has channel mapping, where the app interface needs to show the highlights of latest conversations from respective team channels.

Currently the graph endpoint which exposes the channel messages is beta and available as protected endpoints.

Note: Microsoft Team restricts more sensitive data, and integrate the endpoints if there is a need. Please read out the following notes for the integration guidance.

• This article little focuses on information about integrating protected APIs. As of today, Microsoft has restricted integrating some of their beta APIs. The protected endpoint list is shown here. https://docs.microsoft.com/en-us/graph/teams-protected-apis. 
• To enable the protected APIs for integration, you would need to reach out to Microsoft team for integrations. The request could be submitted through this form. https://aka.ms/teamsgraph/requestaccess. 
• The request is generally reviewed, approved and enabled over a timeline mentioned in the article. 
• The request form needs to be filled with tenant, Azure AD app (explained below) and other details. 
• Once the request is approved, you will get a notification over email confirming the API enablement within specific timeline. 

Azure AD App Registration and Configuration

Before submitting the request, an Azure AD app needs to be registered on the tenant. The use case requires to get the channel conversations. So, the following snapshot shows the configured and granted permissions. Apart from this configuration, the other key parameters needs to be configured.

• Enabling oAuth implicit flow. 
• Enabling implicit tokens 
• And providing redirect APIs. 

API permissions required for integration

Search for Documents from Microsoft Teams Channel Conversations

This article illustrates a sample for finding the documents available on a team, with keyword search from channel conversations. This is achieved with the help of outgoing webhook service, where documents from Team are being pulled with the help of Microsoft Graph API.
#Azure #GraphAPI #MicrosoftTeams #Office365

The following screenshot shows the list of documents retrieved sending the keyword via the channel conversation.

Finding documents from channel using outgoing webhooks