In our previous blog, we delved into the intricacies of SaaS discovery integrations and their pivotal role in comprehending and managing your digital ecosystem. Before we conclude, it's imperative to shed light on a vital aspect of the process: manual intervention in service matching and how it can bolster our machine learning (ML) model.
Client Responsibility in Elevating Service Matching
When it comes to refining service matches, the onus squarely rests on the client's shoulders. Services often span multiple environments like UAT, Dev, and Prod, each adorned with distinct IDs and a fluctuating number of users. This underscores the client's pivotal role in meticulously reviewing these matches. Here are some pivotal considerations:
- Vendor-Service Cohesion: Clients must conduct a comprehensive review to determine the appropriate vendor to associate with each service. This decision carries profound significance from a machine learning standpoint. Best practices encompass scrutinizing services with a discerning eye. If, for instance, a test environment of a particular service is erroneously matched as production, clients should step in to rectify this discrepancy. This proactive engagement not only rectifies inaccuracies but also fine-tunes our ML model. If clients are unsure about which service to match and are connected to Azure AD, there is a convenient way to verify login events. You can access the Sign-in logs under the monitoring section of your Active Directory or perform a search. For your convenience, we have included an attachment that provides a visual representation to aid in your understanding of this process.
Verify service before matching - Comprehensive Service Matching: Consider the practice of matching all related services to one principal service. For instance, instead of matching just 'Microsoft Teams prod' or 'Microsoft Teams test,' consider matching both to 'Microsoft Teams.' This strategic approach mitigates the risk of overlooking any associated services and, crucially, enriches the understanding of our machine learning model. Over time, this empowers our system with greater intelligence.
Comprehensive Service Matching
Displaying Active Users in EAM
Our Enterprise Architecture Management (EAM) tool furnishes clients with insights into active users for specific services in the Fact Sheet. All this information is coming from SMP (SaaS Management Platform). However, the calculation of this metric hinges on several factors:
- One SSO Integration: In cases where a single SSO integration is in play, we present data exclusively from that particular source.
- One Manage Integration: Similarly, if just one Manage Integration is present, we present data from that source as the primary reference. It’s important to note that Manage Integration is only available in SMP for now.
- One SSO + One Manage Integration: When both SSO and Manage Integrations coexist, we give precedence to the data emanating from the Manage Integration.
- Multiple Manage Integrations: In scenarios featuring multiple Manage Integrations, we perform data aggregation across all Manage Integrations. While this approach may occasionally lead to the counting of users multiple times, it's important to note that it also captures instances where a person uses two accounts, thus occupying two seats.
Understanding the mechanics of how active user data is computed empowers clients with a more nuanced comprehension of user activity within their services.
In conclusion, manual intervention and assiduous involvement in service matching play a pivotal role in the success of SaaS discovery integrations. Clients' active participation not only serves to refine our model but also ensures a more precise representation of their digital ecosystem. Should queries arise or should you require assistance, do not hesitate to reach out to our support team. We are steadfast in our commitment to helping you leverage the full potential of SaaS discovery integrations for your organization's benefit.
