Winning with Sisense and Redshift

[pbsisense]

The Power of Partnership With AWS

Sisense and Amazon are partners, which means Sisense can create compelling value for our customers, by delivering services that take advantage of the AWS technologies to solve important problems that they face.

Here, you’ll find recommendations that optimize the value of the Sisense-AWS integration when our customers subscribe to AWS and connect to Amazon Redshift.

Why Work With Amazon Redshift?

Amazon Redshift is a highly scalable and versatile cloud data warehouse. It allows for exabytes of data to be stored and processed and integrates well with the AWS family of products, including S3, EMR, Glue, Athena, SageMaker, among others. Sisense offers connectivity to Amazon Redshift. With Sisense Notebooks, data analysts have at their disposal, a whole new bag of tools to deliver powerful and meaningful insights to business customers. The following are key features that Amazon Redshift users may avail of directly through their admin consoles, or programmatically through Notebooks.

AWS Lambda User Defined Functions (UDFs)

Lambda UDFs make calls to server-less deployments of existing Lambda functions implemented in procedural languages such as Python, Java, among others. In a sense, it is like making a remote procedure call from within a database environment to an external body of business logic. The advantage of using Lambda UDFs in Sisense Notebooks is that analysts may leverage existing procedural assets for performing additional processing that is otherwise difficult to conduct in SQL. For example, consider a data set to be used in solving classification problems. The data contains null entries for some key dimensions. A common technique would be to impute data where there are null entries. There are scores of algorithms available for imputation, but say the analyst wants to use the KNN algorithm to impute the data. A Lambda UDF call would facilitate this task quickly, cheaply, and very effectively. Procedural languages can not only implement complex algorithms, but also call other components, perform exception handling, and error recovery, making them robust. 

Another big advantage of using Lambda UDFs is that a separate compute instance does not require to be spun off. In some traditional deployments, all code resides on a dedicated compute-instance running a server that caters to calls made by external clients. With Lambda UDFs there will be no such administrative overhead. 

Amazon EMR

Amazon EMR scales execution by orders of magnitude on very large data sets. Consider, for example, an analyst who wants to aggregate a billion records. Applying Amazon EMR will parallelize the execution and aggregation to deliver results in seconds instead of minutes. Sisense can be set up for Amazon Redshift and Amazon EMR to allow customers to achieve scale and performance. 

Amazon Machine Learning (Redshift ML)

Redshift ML is game-changing Machine Learning and prediction technology that does not require sophisticated AI/ML skills. It may be invoked with commands in Amazon Redshift to create and use Machine Learning models for generating predictions.  Behind the scenes, Redshift ML interfaces with Amazon SageMaker Autopilot, a model identification tool that takes in training and production data and finds the best model to solve the problem. Redshift ML, if used extensively, will vastly improve the quality of analytics that code-first analysts can deliver to business users.

By calling Redshift ML from Sisense Notebooks, analysts will be able to infuse AI in business and mobile applications. For example, if a manager wants to estimate sales for the next month, a forecasting or a regression model, or for that matter any other model, whichever makes the best fit, will automatically be selected and the results rendered through the Sisense Infusion framework. Sisense Notebooks with Redshift ML will empower analysts to new heights and capabilities.

Federated Queries

Federated Queries allow Amazon Redshift to bring data from external PostgreSQL and MySQL data sources.   This feature reduces the need for ETL processes and vastly expands the reach of Amazon Redshift. Customers who prefer to use Amazon Redshift for aggregation and rendering but want to store raw data in a cheaper and slower MySQL or PostgreSQL databases, may use federated queries to facilitate bringing selected data into Amazon Redshift, and on an as needed basis. Federated queries save time and money for the administrator as well as the analyst. 

Amazon Redshift Serverless

This feature allows Sisense to start Amazon Redshift instances without the need for any of the cluster management and other administrative tasks. The deployment can scale like any Amazon Redshift instance.

Benefits

Connecting your data to Amazon Redshift with Sisense gives you access to this fast and powerful, fully-managed, petabyte-scale data warehouse service in the Cloud. This connection enables you to store data in the Amazon Redshift database, and query the data in Sisense. As a result, you can enhance the performance of your BI and analytics, improve cost control and deliver greater ease of use that optimizes your Cloud investments.

Pat Bhatt is Director of Product Management, Cloud Analytics, at Sisense. He has over 20 years of experience in product management and innovation in the tech space, having led product management at Model N, SkyNovus, Intuit, and Silicon Valley Bank.