Summary

HP Industrial Print modernized its data platform by moving from a siloed, rigid architecture to the Databricks Data Intelligence Platform, enabling faster onboarding, improved governance, and seamless data sharing with customers who produced data through app interactions. This transformation boosted pipeline performance by 40%, and unlocked new revenue opportunities through scalable data products and monetization.

HP’s Industrial Print Software Solutions (IPSS) Business Unit has always stood at the intersection of cutting-edge hardware and software. Their portfolio provides sophisticated software and analytical products, offering digital presses workflow, monitoring and analytics. But as demand for high-speed, flexible, and automated printing grew, so did the need for a more intelligent and scalable data platform. Though robust, HP’s legacy data infrastructure limited its ability to move fast, collaborate broadly, and fully capitalize on its data. That’s why HP turned to Databricks.

The Role of Data in HP Industrial Print

To understand the importance of this transformation, it’s worth looking at how data flows within HP Industrial Print. When customers place print orders, everything from custom packaging to wide-format graphics, HP routes these requests through its proprietary application, PrintOS Site Flow. This system connects the customer with one of HP’s global network of Print Service Providers (PSPs), who fulfill the order. As the job progresses from onboarding to printing, packaging, and shipping, the PSPs scan barcodes and update statuses, creating a rich stream of operational data. This data includes orders, provider assignments, material specs, and timestamps.

From this foundation, HP extracts insights to drive business decisions. Dashboards help PSPs manage workloads and performance. Internal analytics teams use the data to monitor customer engagement, optimize supply chains, and ensure billing accuracy. HP also empowers its partners by exposing this data so PSPs can run their own comprehensive analytics.

In short, data is both an operational backbone and a strategic asset for HP Industrial Print. But the systems powering it weren’t keeping up.

The Challenges of the Legacy Architecture

In the previous setup, data flowed from MongoDB through a Kubernetes-based pipeline running on Amazon EKS. Transformed datasets landed in Amazon Redshift for internal analysis and in Amazon RDS to serve external applications. While functional, the architecture came with trade-offs.

Sharing data across HP business units was complicated and time-consuming, often requiring custom pipelines or manual data exports. The lack of a medallion architecture meant it was challenging to trace data lineage or reprocess historical data when logic or business rules changed. Governance was handled in silos, leading to inconsistent access policies.

Perhaps most critically, this architecture stifled innovation. HP had ideas for new data products—services combining internal and external data to deliver deeper insights or generate revenue—but lacked the agility and visibility to implement them.