Development Tips

BI Security: Best Practices when Embedding Analytics in Applications

By Michelle Gardner
Share on LinkedIn Tweet about this on Twitter Share on Facebook

For application teams, BI security scenarios can be complex and have very precise requirements. Considerations may include partitioning multi-tenant data, limiting access to individual records in the data, or securing different parts of the application, workflow, and capabilities based on individuals’ rights and roles.

Implementing these scenarios in an analytic application can be a daunting task, often requiring redoing work that’s already been done setting up the application’s existing security model. If software engineers have to recreate or replicate authentication and authorization information in their analytics tool, it can negatively impact their ability to maintain, grow, and adapt the product over time.

>> Top Requirements for Business Intelligence Projects: Get the BI Buyer’s Guide <<

Fortunately, some embedded BI vendors are finding new ways to avoid this rework. By seamlessly leveraging your existing security model in your preferred analytics solution, you can save days or weeks of effort while also managing all your security settings in one place.

Why Preserve Your Existing Security Model?

Most companies have invested significant time and effort in their security models. They’ve carefully designed them to allow the right parties access to the right information throughout their application. But traditionally, few analytics vendors have leveraged those existing security models—resulting in an inefficient structure that stores user information across multiple systems.

The risks of the traditional approach to BI security include:

Wasted developer time. Instead of iterating on the core IP, developers are forced to address BI security problems caused by replication or mapping.
Poor user experience. If security is not replicated on the right schedule, it could lead to limited access for users, potential BI security risks, and an overall poor experience.
Limited scalability and flexibility. As you make changes to your security permissions and to the data model, it could require reimplementation and limit sharing possibilities between groups with different access levels.
Delayed release. Replicating, testing, and deploying a duplicate BI security model requires time and testing resources that will inevitably delay product releases or updates.
No support for multi-tenant deployments. Many analytics vendors build proprietary BI security systems that do not support multi-tenant deployments or the sophisticated security requirements of analytic applications.


The Value of Adaptive Security

Why re-invent the wheel when you don’t have to? By leveraging an adaptive BI security model and token-based API, you can effectively eliminate all of the above risks and save yourself time and effort.

The benefits of the adaptive approach to BI security include:

• Reuse of existing user roles and rights in multi-tenant architectures
• Managing security in one place rather than once in your application and again in your analytics solution
• Data security at the row, column, and table level
• Easily updated since you can make changes to user roles, rights, or access changes all in one place
• Support for multi-tenant environments through passing of important information—like which database to use for the current customer

An adaptive security approach to BI provides the best of both worlds: seamless management of users in your existing application and the ability to pass all your existing rights, roles, and other security information to the new application.


More on BI Security: 


Originally published June 1, 2017; updated on August 10th, 2017

About the Author

Michelle Gardner is the Director of Corporate Marketing & Communications at Logi Analytics. She has over a decade of experience writing and editing content, with a specialty in software and technology.