Choosing the Right Embedding Technology
This video will give you an idea of what features you should look for when choosing an embedded analytics technology.
First, of course, make sure it’s designed to be embedded. A lot of BI vendors out there peddle legacy technologies that have been retooled for embedding analytics -- with very mixed results.
Some other important capabilities include speed. Your embedded analytics technology should be fast. Users don’t want to watch a spinning hourglass while they wait minutes for query results. It should also natively accommodate modern data sources and make it easy to add new sources. When your embedded technology has these and a few other essentials, it can keep up with the analytical demands of big data.
Here's some things to consider when you're looking for analytics technology to embed. First of all, of course, you want to make sure that the technology has been designed to be embedded, and you want to make sure it provides the APIs, the SDKs you need to really make it look and feel like part of your application.
Is It Fast?
Secondly, you want to make sure it's fast. We're all dealing with more and more data these days. You hear more and more about big data where people have, potentially hundreds of millions or billions of rows. But, as humans, we want response times of under five seconds, ideally under one second. So, you want to make sure your analytic technology is gonna meet those performance goals.
Can It Handle Modern Data Sources?
You want to make sure it's gonna be able to adapt to modern emersion data sources. You want to make sure your analytics technology is not locking you into, yesterday's technology and, just making sure it can adopt as these new data sources come out.
Can It Analyze Streaming Data?
You want to make sure it's appropriate for real time streaming sources if that's part of your application - so, the ability to consume this real time streaming data, to update the visualizations and dashboards in real time and to also blend that real time data with historical data so you can look at, comparing today to yesterday to last year.
What About Security?
Security is another important consideration. So, you want to make sure it supports things like single sign on for authentication. You want to make sure it has very detailed authorization capabilities typically at the row and the column level.
And then also, you want to make sure it supports things like delegated security. More and more, we see data sources being controlled or data access being controlled at the data source level - so, the ability for the analytic application to leverage that data, that authorization capability within the data source.
Does It Offer Deployment Flexibility?
You also want to make sure the analytic technology provides you the deployment flexibility you need, depending on how your application is deployed. That may be on premise, it may be in the cloud, it may be as a software as a service application, or it may be a hybrid combination of any of those.
Can It Scale?
And then finally, you want to make sure that it's gonna support the kind of user load you have in mind. So, traditional software applications were delivered as a single monolithic server. The more much more modern approach is around micro services where you break down that server into the little tiny micro services, which are very easy to distribute across multiple servers, either on premise or in the cloud. And this can enable you to support a cloud scale deployment of potentially tens of thousands or hundreds of thousands of users.