GRAX History Stream

- Welcome to the next evolution of GRAX-- History Stream. History Stream is the latest data-ops innovation that helps you unleash the value of your historical Salesforce application data. It's always been GRAX's mission to protect your data and the value of your data. That's why GRAX was built on the pillars of data ownership, data access, and data reuse. By owning your data in your own cloud environment and retaining access to all of your historical data sets, you're now able to reuse your history downstream.

Did you know that your history starts at backup? Once you start capturing your data over time with GRAX, you're actually feeding your history into your GRAX data vault. Let's execute a customer account backup now.

Once securely stored, you can see your historical versions of data stored as parquet files in your own object bucket such as your S3 bucket. These versions of your history will continue to be automatically captured each hour. History Stream was built to align tightly with how data processing pipelines work in the modern enterprise. And thus GRAX leans on well-established and understood formats, frameworks, and conventions.

With that in mind, let's take a look at reusing this data set to make machine learning predictions easily accessible. What better way to strategically leverage your history than creating, training, and deploying machine learning models with a few clicks. In this demo, we'll feed our history into SageMaker, the AWS cloud machine learning platform. Let's take a closer look. We'll use the real-world, publicly available, anonymized data set that shows customer account data. We backed up our customer data set using GRAX, as you saw. We've turned on History Stream, and we've sent the data to Redshift. And now we'll leverage Redshift ML to create our SageMaker training jobs and inferences.

First, as you can see here, we'll create an ML model that predicts churn based on all the other data points on the Salesforce object. Next, let's ensure the model was created. As you can see here, the model is ready, and we can see details about this SageMaker ML model. Then we can run simple SQL queries that use this predict churn model that we created to see the actual predictions in the SQL query results, as you can see here. Finally, we'll create a view that captures all of this information. We can see in AWS SageMaker that training jobs and inferences have been created.

Now with a single click, we can get all of this data and inference predictions into AWS QuickSight to make it valuable and actionable for business users. We can see things like total revenue that will churn, number of churning customers by contract type, by tenure, and by monthly spending habits. Additionally, with just a few more clicks, we did the same process to create a separate model that predicts total customer revenue and not just churn. And we can just as easily create QuickSight reports and dashboards that show customer revenue prediction patterns based on the GRAX history.

We can see things like revenue predictions by payment method, by contract length and status, as well as by monthly spending habits and tech support statuses. As you can see, backing up and archiving my Salesforce CRM data at the highest fidelity, I can leverage this history to reuse and answer my business questions and make key business inferences and predictions confidently and simply. Ready to unleash your Salesforce data's value? Just click the Get Started button in the upper right-hand corner to get in touch, or email sales@grax.com.