I am at AWS re:Invent 2017 this week, and I attended a session on Amazon QuickSight. QuickSight was always something that I wanted to try and learn more about, and the session really helped me get a good initial understanding of what it is and how people can use it for business analytics. To introduce, Amazon QuickSight is a business analytics service that you can use to gather data, perform analysis and generate custom dashboards to get business insights from your data.
Let’s walk through the workflow on using QuickSight and talk about the different components while doing so:
Step 1: Data Sources
Data is everywhere! You are probably collecting all kinds of data in your environment, from user data to metadata to system logs. If you are already using AWS, you are probably uploading all this data unto AWS S3 buckets. But, in a majority of the cases, you do not know how to use this data to get better insights and improve your business. Amazon QuickSight can help you with that. It supports a whole plethora of data sources ranging from Relational Data sources like Amazon S3, Amazon Redshift, Presto, Snowflake to File and SaaS Data Sources like .xlsx, .csv. and Salesforce.
Step 2: Data Preparation
That’s great, QuickSight supports all these data sources, but what next?? Next step is Data Preparation, where you clean and transform this raw data so it can be used with Amazon QuickSight. In this process, you will create what we call Data Sets. A data set identifies the specific data in a data source that you want to use. You can also use data preparation to join tables or use custom queries to fetch the data you want from your data sources. For eg. If you have a SQL database for your Twitter stream, then you can write queries to only fetch the tweets from the North America region. This can be your data set that you want to analyze in the next step. Use this link to learn how you can create these Data Sets from your Data Sources: http://docs.aws.amazon.com/quicksight/latest/user/creating-data-sets.html
Cool, now that you have a data set, you are ready to create an analysis based on that.
Step 3: Analyses
An Analysis is the baseline unit for creating and interacting with visuals and graphic representations of your data. Think of Analysis as a container for a set of related visuals and stories. For eg. All the key performance indicators that you want to highlight. Let’s say you want to analyze storage performance. One visual can be a scatter graph for IOPS, another visual can be a line graph of the throughput, etc. All these visuals combined are what form an Analysis.
Step 4: Dashboards
After step 3, you have an Analysis that you can share with your team or the larger organization. You can either choose to share the Analysis as it is, or you can create and share custom dashboards in Amazon QuickSight, which are nothing but a read-only snapshot of your analysis that they can use to get business insights from your data. Sharing Analysis or Dashboards each have their own pros and cons. By sharing Analysis, you enable collaboration between your team, where other people can use the same data set and add/edit the visuals to your Analysis. And if you want to just share the final data, then you can share the read-only Dashboards.
The following is an image from AWS’s blog. Here you see an Analysis view, where you can add visuals and once you are done, you can click the share button on the top right to create a Dashboard:
Hopefully, this blog was helpful to get a basic understanding on what Amazon QuickSight is and what the different components are. I didn’t get a chance to cover SPICE, which is QuickSight’s Super-fast, Parallel, In-memory calculation engine. It helps perform those advanced calculations and serve data. You can learn more about SPICE using the following link: