With an important update to the Salesforce-Snowflake partnership, Einstein Analytics and Snowflake can deliver even more value to customers. Integrations between the two will allow organizations to combine and analyze all data in Snowflake Cloud Data Platform and visualize it and derive new insights with business intelligence tools such as Tableau and Salesforce applications.
Since announcing our own partnership with Snowflake recently and diving into the highlights, we spoke with our team about the benefits of the Einstein-Snowflake partnership and what stakeholders can expect.
Dave Dixon, Analytics Practice Lead & Director of Analytics
Snowflake is a huge accelerator for insights delivery in Salesforce, Einstein, Tableau, and other systems. Two new features I’m especially excited about are:
- The ability to write your Einstein data directly into Snowflake (pilot feature in summer ‘20)
- The ability to directly query Snowflake in real-time from Einstein (beta feature in summer ‘20)
Using Snowflake as Your Core Datastore for Einstein
The ability to use Snowflake as your core datastore for Einstein is a game-changer. Many organizations have objections to copying data to multiple reporting platforms, and frankly, there is some risk with that approach; data can get out of sync, resulting in multiple versions of the truth, etc. With these new Einstein/Snowflake query and integration capabilities, organizations can use Snowflake as the central hub for their enterprise data, and surface it in Einstein, Tableau, or whatever system makes sense for the use case, and user experience. The data doesn’t have to be copied to local Einstein stores. It can be queried directly from Snowflake.
Additionally, predictions can potentially be written from Einstein into Snowflake and then consumed by users outside of traditional CRM roles, such as supply chain, logistics, or HR users. In essence, Snowflake becomes the unifying data architecture that allows your CRM, prediction engines, enterprise data sources, and analytics tools to all play nicely together.
Dustin Weaver, Snowflake Practice Lead and Analytics Architect
Organizations making the switch from traditional data warehouses to Snowflake will eventually ask the question: How do I use this data to help build predictive models?
The great thing about Snowflake is that it enables your organization to bring all your data into a centralized, easy-to-access database. Having data from all these various sources can enable an organization to gather better insights and make fact-based decisions on their future. As part of that future decision-making process, tools like machine learning can assist an organization’s people in learning which parts of their data have the most impact on those decisions. Traditional machine learning tools can provide those insights, but have steep and complex setup and configurations. On top of that, many of these tools can’t provide their output back to Snowflake. Having predictions sent back to Snowflake enables organizations to surface these predictions across business units and provide results in the tools already in use by various departments.
Creating Predictive Models in Einstein Discovery and Surfacing the Results in Your Snowflake Data
Salesforce’s Einstein Discovery product enables citizen scientists within your organization to quickly and easily create machine learning models. Using Salesforce’s Einstein Prediction Service Scoring API, you can now create predictive models in Einstein Discovery and surface the results of these models in your Snowflake data.
Kyle Bowerman, Data Science Architect
Snowflake is the cost-effective data workhorse that we have all been waiting for. It provides the best of both worlds by providing the structure of data warehouses and the integrity and flexibility of a data lake.
Additionally, they solve the monolithic availability issues by providing a true distributed mesh on the world’s most trusted cloud infrastructure. With their new direct data connector to Einstein, Snowflake is enabling companies to become data-driven organizations by allowing them to derive fast, flexible, powerful analytics and insights from deep inside their core without worrying about digital plumbing.
Einstein and Snowflake Support from Atrium
As market leaders in building intelligent systems, we can help you get started with Einstein and Snowflake or extend the benefits of the two. Whether you need help building a predictive model, surfacing the results in Snowflake, or understanding the right usage for you, our team is here to support your data needs.
Contact us to see how we can help power your data initiatives with Einstein and Snowflake.