What is our primary use case?
I'm a cloud consultant and prefer Azure Purview as a governance tool. A few of my clients are in the banking sector, including global and Indian banks. They need a centralized solution for compliance and governance.
I haven't used Purview much for SaaS or device management, but I have explored the SaaS services and PaaS in terms of other tools that can be integrated with Azure Purview. Our clients often have an ETL solution that's in Azure directly. Most of our clients use Synapse Analytics or Databricks. One of our energy clients in Australia has the same use case for data platform implementation. They have some unstructured data on-premises, and it goes back to the Azure storage account as raw data. They do some transformation with the help of an ETL tool. For agentless forecasting, they do model training with the help of Databricks or Azure machine learning. They'll use Power BI for visualization, but Purview is the centralized governance tool.
How has it helped my organization?
Purview improves compliance and governance with data source and auto-scanning features. All those data assets would be in a centralized repo that pulls from different sources and databases that hold the bank data or the customer's data. The solution enhances compliance in terms of security factors and issues like duplication. It scans for new data coming in and older data that already exists. Purview collects the metadata to get a clearer visualization of all the compliance aspects.
Another benefit is accelerated migration. Purview facilitates smooth migrations by giving clients insights into their data, which data is useful, and what kind of data is non-compliant, so you can classify data based on compliance, priority, and utility. It helps clients decide which data is essential and should be migrated. Identifying non-compliant data can help clients improve security and privacy.
Data analysis and classification isn't a two-day or two-week job. It's a long-term process. It might take two or three months to gather feedback on what kind of data is the most beneficial, like structural data, static data, etc. The data lineage doesn't come in one or two days. It takes some time to get a complete picture of your source and destination and the data cycle from start to finish. You can say that the data was in this phase six months ago, and now it is in this phase. Six months is enough to get a clear picture of the data flow and the kinds of data that are most beneficial. Purview is helpful for long-term data management and classification.
Purview improves visibility. If I go back to a year ago, when I was learning Purview, my perspective was different. There were fewer features, and Purview has grown at a rapid pace. I was originally a network engineer, but I am now a cloud consultant, so it was challenging to work with the Azure version initially. I started seeing the benefits of its data classification features once I started consulting, seeing everything that comes into a client's bucket, and receiving feedback. Now, I can see the solution's strengths in governance and compliance.
The auto-scanning functionality and automation features in Azure have a positive impact when connecting data sources. When there are multiple data sources, we can make connections and start discovery. Manual discovery will take so long, and we don't know when the new data is coming in, so automation is useful. When new data comes into the databases, Purview triggers that auto-discovery part so the data stays updated. It's more efficient and more accurate.
For example, we have let's say we have data coming in this month. We have a discovery scan, and some new data comes into the databases after a day or two. If we don't run the auto-scanning feature, the data will get outdated. It would affect if we need to present something or calculate something. We can discover data quickly and get accurate data. It isn't completely in real-time. It still takes time to refresh things, but we haven't
Automated discovery reduces the amount of time needed to take action on insider threats. If discovery is already done, it's easier to classify the data and import it into a data visualization tool like Power BI. We need to complete the discovery before moving on to the data lineage component or the Data Factory pipeline. It reduces the time by about 12 to 15 percent.
What is most valuable?
My favorite Purview feature is auto-scanning. Once we set up Purview, we can automatically scan multiple data sources when new data comes into specific databases, like SQL and Oracle. We don't need to rediscover the new data or do anything manually because it automatically happens.
Purview also offers some additional integration capabilities if you use the Azure edition. We can seamlessly integrate tools like Azure Data Factory and Synapse Analytics to provide analytics and data transformation services in a customer's ETL pipeline. Organizations want to structure that data, so they use a basic ETL tool, which is commonly Azure Data Factory. Purview provides us with all the connectors needed to integrate these data tools. Another feature I like is data lineage. Purview tracks the data from its source to the destination.
I haven't seen many challenges with integrating or supporting native Microsoft solutions like Office 365. We haven't come across anything in Microsoft 365 that's unsupported out of compliance with HIPAA regulations in the health sector or banking and finance regulations. It conforms to PCI DSS compliance methodology or GDPR.
We have multiple clients in the energy and banking sectors. Purview is vital in data platform implementation projects involving ETL transformations and model training. It's our default tool for governance when we're pitching our organization during presales. However, our smaller customers often don't need that much governance, or they're good with Azure Monitor. They may also prefer some other governance tool or might have an on-premise tool that they are already using. They don't want to change it despite the integration and features. Our big customers may have existing governance tools, but they want to use Purview because it offers additional features. It provides them security, compliance, and the flexibility to integrate with third-party and Azure native tools.
What needs improvement?
Two features are unsupported—custom insights and the DLP component—that would be beneficial to me as a consultant and for the customer in terms of security and monitoring. Regarding security, DLP would provide a more granular level of data masking. Custom insights would offer more detailed monitoring and alerts that can notify customers of failures or anything requiring urgent action.
DLP is not a part of Purview. Our larger customers require some advanced features, such as dynamic data masking, encryption, and decryption. For example, some of our projects in Dubai involve machine learning use cases and encrypted critical data on-premise. It varies. Data encryption and masking are not priorities for some customers.
Microsoft has some built-in data masking tools. Some customers believe that masked data is safe, and they don't want to move it. We tell the clients that Purview doesn't move the actual data, only the metadata. The customer is convinced that DLP is not part of Purview, but that is not a concern because it's all about metadata. The original data is not transferred from on-premise to Azure. Purview is not storing the actual data. It takes the data to perform discovery and provide better data classifications. If DLP is added, then Purview will be stronger.
When I talk to clients about these DLP features, they say it still lags behind in data integration and support. It does not affect the sales side or prevent us as consultants from convincing them to switch because of these two unsupported features. However, Purview does need some improvements in data security and third-party integrations.
For how long have I used the solution?
I have worked with Purview for a year.
What do I think about the stability of the solution?
I rate Purview six out of 10 for stability. It is stable, but they're constantly adding new features, so it needs more stability in the future.
What do I think about the scalability of the solution?
Purview is a scalable solution because Microsoft manages everything on the back end. We don't need to handle any databases or servers. We can talk about the general scalability of Azure services, not Purview specifically. The SLA gives an uptime of 99.99 percent. I have not had any scalability issues with Azure services, and this goes for Purview as well.
How are customer service and support?
I rate Microsoft support seven out of 10. They respond quickly and follow up fast after the issue has been closed to see if the problem is resolved.
How would you rate customer service and support?
Which solution did I use previously and why did I switch?
I have not used GCP, but the AWS centralized governance tool lags in terms of authentication and authorization. AWS also has features like data lineage it can provide. However, Purview stands out for security protections and role-based access control. Purview offers better granularity compared to the AWS or GCP governance tools. The access control list enables granular levels of access to its users.
How was the initial setup?
Deploying Purview isn't too complex. The initial installation takes about eight hours. Setting up a data governance solution involves configuring the networking components. The networking part can be done in parallel with deploying other services. However, you must completely implement the data platform before connecting the data sources and performing data discovery. Later, you can begin the data analysis and classification on Purview.
The data deployment and networking configuration is similar to deploying other Azure services. I imagine on-prem Purview requires more time to set up. If we are giving a client an estimate of the time it would take to complete the project, the deployment and networking part would not additional time. However, the data discovery, classification, lineage, and data source connectivity, require two more weeks. The total deployment takes five to eight weeks plus an additional two weeks for the discovery phase.
The networking engineer who deployed Databricks, Azure Data Factory, Synapse Analytics, etc. can easily deploy Purview because it's the same. However, we require a dedicated Purview engineer for the discovery phase. We haven't had to do much maintenance so far.
What's my experience with pricing, setup cost, and licensing?
Microsoft Purview is priced in the middle. It isn't the cheapest, but it isn't the most expensive. It's affordable compared to other public cloud services. Purview costs about 20 percent less than AWS, but it is still expensive compared to other Azure services and governance tools.
What other advice do I have?
I rate Microsoft Purview seven out of 10. I recommend Purview over other governance features because it has multiple features that make it stand out from the rest. Once it is updated with features like DLP and custom insights, Purview will be a market leader. These additional features will help the solution earn more enterprise-scale customers.
Which deployment model are you using for this solution?
Public Cloud
If public cloud, private cloud, or hybrid cloud, which cloud provider do you use?
Microsoft Azure
Disclosure: PeerSpot contacted the reviewer to collect the review and to validate authenticity. The reviewer was referred by the vendor, but the review is not subject to editing or approval by the vendor. The reviewer's company has a business relationship with this vendor other than being a customer: Partner