How to Elevate Your Databricks Lakehouse Into a Composable CDP

Looking to turn your Databricks lakehouse into a customer experience engine? You’re going to need a (composable) boat.

The ActionIQ team just touched down from the Databricks Data + AI Summit in San Francisco, and we’re still buzzing from the big MosaicML acquisition announcement, conversations with fellow data-obsessed professionals and more — but one of our favorite moments was sharing the stage with Northwestern Mutual to see how they turned their lakehouse into a CDP.

At the Databricks Data + AI Summit, Justin DeBrabant, Senior VP of Product at ActionIQ talked with Jeff Parkinson, Vice President of Core Data Engineering at Northwestern Mutual about their architectural evolution and his insights into how they designed a tech stack for the future.

Below, see how the team at Northwestern Mutual went from data lake pioneers to composable trailblazers by designing a flexible stack with Databricks and ActionIQ. Scroll down to read our takeaways:

How Northwestern Mutual Elevated Their Databricks Lakehouse Into a Composable CDP

Northwestern Mutual’s Architectural Evolution

The team at Northwestern Mutual knew something was wrong when they were seeing a proliferation of fragmented data with their on-prem database, with about 200 data labs scattered throughout the organization. They wanted to achieve proper data lineage with all of the classifications of data, and all the security measures — they knew it would be easier to do that and cut down copies by putting it all in one place.

They decided to adopt the data lake and move to Databricks as their central store of data for a centralized data lakehouse with an activation engine — enter ActionIQ.

They had three key priorities when they were redesigning their infrastructure, and found that the combination of Databricks and ActionIQ would meet those criteria in the following ways:

1. Simplify Data Footprint and Reduce Handoffs

Data fragmentation had become a major problem for Northwestern Mutual — when they were designing their stack, they were looking for a solution that would allow them to reduce some of the friction with their data engine and simplify their data footprint and bring their data together securely in one place.

It became evident early on for the team at Northwestern Mutual that pairing Databricks and ActionIQ was perfect for reducing handoffs with a CDP that requires no copies whatsoever — tapping directly into the lakehouse.

2. Ensure Governance and Compliance

Privacy, governance and security are non-negotiable for any organization — but for a company like Northwestern Mutual, making sure that compliance and reduced risk are built into the fabric of their processes is paramount. With a composable CDP, no copies are required to query the data, which eliminates the need to worry about copies on copies of data increasing risk and cost. With a composable model, ActionIQ taps directly into Northwestern Mutual’s secure, central store in Databricks to activate data.

For Parkinson, the benefit was evident. He said, “Governance and privacy, and an infosec standpoint, saves so much time when you’re solving one area — that’s the golden standard that different companies are working towards, and it’s key for you to find that speed to market.”

3. Streamline Operations and Speed-to-Market

Before Databricks and ActionIQ, the team at Northwestern Mutual took months to get a single marketing campaign out the door, which often meant that the campaign was no longer relevant. With ActionIQ’s HybridCompute, business teams can access data directly and quickly, with models that are ready-to-be-utilized for campaigns like cross-sells, upsells and everything between acting on the freshest data available from Databricks.

Not only do the marketing teams enjoy the simplicity of activating their campaigns without waiting months for a request, but data engineers enjoy the simplicity of having their data centralized in the lakehouse — without having to deal with backlogs of requests.

Hybrid Architecture Offers Northwestern Mutual the Flexibility to Move Fast

ActionIQ’s approach balances the practicality that many enterprise organizations will likely have data outside of one single system, even if the goal is to get it all in the lakehouse with the lakehouse as the center of gravity, organizations get a flexible approach with pre-built integrations. The hybrid composable strategy brings it all together in the most flexible way possible — like a choose your own adventure. These lines will change over time, DeBrabant explains, so a flexible approach offers a way for brands to maintain that speed and evolution.

For Northwestern Mutual, over the last few years of evolving their tech stack with Databricks and ActionIQ, Northwestern Mutual has increased their flexibility and speed to be ready for anything — without wasting any downtime.

Composable CDPs require no copies to activate customer experiences, so organizations save cost. DeBrabant explains, “We all know that moving data is expensive, and not just expensive from a compute standpoint. From an operational standpoint, data pipelines, always changing and evolving. People have their own data, they change the format, something changes, it breaks, you need to go investigate, it’s very operationally expensive to maintain. You don’t need to do this.”

Turn Your Databricks Lakehouse Into a CDP With ActionIQ

With a hybrid CDP that seamlessly works with their data lakehouse, Northwestern Mutual is looking ahead with confidence at a future full of change (and generative AI). With a clean, centralized data source and a composable CDP, they’re poised to explore, test and learn in a flexible future.

To learn about how your team can take advantage of a composable CDP with Databricks, check out our ActionIQ x Databricks Solution Sheet to learn how we maximize tech investments together.