Product descriptions:
As the global leader in personal computing and printing technology, HP processes vast amounts of data from millions of devices, services and applications. With Databricks Assistant, AI/BI Genie and Mosaic AI, the teams see the potential of generating faster, more accessible analytics. They are eager to explore Databricks capabilities to accelerate analytics workflows and unlock AI-driven forecasting and strategic decision-making in the near future.
Enabling faster analytics across the enterprise with AI/BI Genie
Various teams across HP’s enterprise are leveraging Databricks AI/BI Genie to make data and insights more accessible, informing better decision-making.
HP’s Analytics and Insights team, part of the Technology and Innovation Organization (TIO), focuses on leveraging data to improve business outcomes for the print business. One of their key use cases centers around understanding supply and demand dynamics to optimize channel and customer inventory levels and minimize over- or underselling, challenges that directly impact GTM efficiency, channel inventory and overall profitability. As Michel Macler, Sr. Data Science Manager in the TIO, explained, “The complexity is daunting. With thousands of SKUs across many countries, anything we can do to sort out anomalies or perform mix analysis quickly would be extremely valuable.” These challenges became particularly acute during the pandemic, as supply chain disruptions made forecasting and inventory management more difficult.
Michel’s team recognized that traditional approaches to diagnosing and addressing sales and inventory issues were time-consuming and labor-intensive, often requiring significant manual effort and expertise. Lisa Wolfe, a Principal Data Analyst, noted, “We had to write long SQL statements to get the answers, and it took time writing them the exact right way.”
Bruce Hillsberg, TIO’s VP of Data Engineering and Insights, echoed these challenges at a strategic level. “We have situations where questions come in during a CEO meeting — like, ‘What’s our market share on ink and laser products?’ — and the idea of submitting a ticket and waiting for someone to build a dashboard just isn’t practical.” This inefficiency underscored the broader challenge: Business stakeholders across HP needed a faster, more intuitive way to access data immediately rather than waiting days or weeks for a new dashboard.
TIO also relies on data to drive strategic decision-making for their product management teams — from understanding customer behavior to assessing feature adoption opportunities. The company collects vast amounts of data from hundreds of millions of devices, including printers and PCs, generating trillions of events annually. Leveraging this data effectively required a level of technical expertise that many of these product managers did not have.
Product teams track key product metrics such as feature usage, battery life and customer interactions, but accessing and analyzing this data was traditionally an inefficient and fragmented process. Product managers would request reports from analysts, creating a dependency that slowed down decision-making. As Bruce explained, “We want the product managers to be able to use data in their decision-making, rather than relying on anecdotal feedback or the last customer they spoke to. We want to get to a mode where the product managers themselves can ask a question to the system rather than submitting a ticket and waiting for an analyst to pull the data.”
Meanwhile, HP’s digital commerce business relies on data-driven decision-making to optimize marketing performance, drive sales and enhance customer experiences. Laurent Laguerre, HP’s Global Head of Data Monetization, recognized the need for a more agile, scalable approach that would empower business users with self-service analytics — without requiring them to rely on engineering teams for every request.
HP’s Digital GTM organization supports a vast ecosystem of data across e-commerce, CRM, marketing and merchandising. The team manages critical business functions, including marketing attribution, media mix modeling, customer engagement analytics and product performance tracking. These insights drive HP’s global marketing and sales strategies, but traditional BI tools — such as Tableau and Power BI — were becoming bottlenecks rather than enablers. Complex queries had to be routed through engineering teams, slowing down time to insight. “We didn’t want our stakeholders to be constrained by the static structure of our dashboards or the reports we generated,” Laurent said. “They needed the ability to ask their own questions in real time, without needing to know how to code in Python and SQL or how to put something together in Tableau.”
During key sales periods like Black Friday and Cyber Monday, there are valuable opportunities to empower marketing and pricing teams with near real-time insights, enabling them to respond proactively to competitive moves. As Laurent emphasized, “When you’re adjusting campaigns and promotions in a fast-moving market and need to know within half a day if it’s working or not, you can’t afford to wait for answers. By the time we were getting these marketers the data, it was too late.”
As HP’s marketing teams shifted toward AI-powered predictive analytics, many business users lacked the technical expertise to leverage machine learning models effectively. These GTM teams needed a way to abstract the complexity of AI-driven insights and make forecasting and optimization accessible to nontechnical users.
The growing need for accessibility and speed of analytics led all these teams to adopt Genie for product management and marketing use cases. “The team spent a lot of time and money building a custom text-to-SQL solution, but now they’ve abandoned it in favor of Genie,” Bruce added. “We just realized that it’s smarter to partner with people who have certain expertise that we don’t so that we can focus on what we do best.”
By integrating various data analytics and programming capabilities into a single platform, Genie eliminates the inefficiencies caused by fragmented tools. Teams can leverage Genie to simplify the identification of anomalies and data issues. Instead of manually building queries from scratch, Genie’s natural language processing capabilities can help users ask diagnostic questions in plain English.
Paul Hammon, Sr. Data Science Manager in the TIO, echoes these sentiments: “Even as someone comfortable with SQL, it can be tedious to look up field names, refine queries and validate results manually. Genie eliminates those pain points by allowing us to ask natural language questions and get accurate, usable outputs in seconds.”
For example, Lisa could use Genie to identify outliers in the Latin America printer market between October and November 2024. She might ask, “Find the top 10 most extreme outliers in the Latin America market.” Genie would then filter and rank the printers by outliers in usage patterns that may help identify potential data issues or unexpected user behavior.
At the same time, Laurent’s team transformed the way marketing, merchandising and pricing interact with data, shifting from static reporting to a fully self-service analytics model powered by Genie. “Instead of waiting on engineers, teams can now ask questions in natural language — right within the tools they use every day, like Teams,” Laurent explained. “They simply click an icon, open a window, type a request and Genie does its magic — instantly translating their query into actionable insights.”
With trillions of data points flowing into Databricks, HP needed strict governance and security controls as Genie adoption scaled. Unity Catalog provides role-based access control (RBAC) and ensures data privacy and compliance across business units as data and answers are shared via Genie and AI/BI Dashboards. By investing in Unity Catalog, HP is safeguarding data privacy while bringing insights to life.
Genie is transforming product management by making product telemetry instantly accessible. “With Genie, our stakeholders can now ask print telemetry trend questions in real time rather than going through a request submission process,” Bruce said.
As for Laurent’s Digital GTM teams, data democratization was the key to unlocking business value. “Genie has the power to disrupt the way we think about analytics. It has the power to enable all the analysts, who must access insights quickly and make better decisions.”
HP is also exploring the integration of AI/BI Dashboards to complement Genie and further reduce reliance on external BI tools. With Databricks’ AI/BI Dashboards, HP can generate data visualizations and reports more efficiently. Additionally, with Genie embedded directly into the dashboards, users can seamlessly ask follow-up questions based on the insights presented in the charts. “We’d love to consolidate to fewer tools, and Genie’s integration with the rest of the Databricks Platform that we already rely on is a big factor in that strategy,” Bruce said.
Unlocking the full potential of Databricks
Looking ahead, HP also sees an opportunity to leverage the full potential of Databricks’ wider technology ecosystem, including Mosaic AI for AI model training and inference. As a leading vendor of PCs and printers, HP wants to integrate AI into their data strategy while ensuring that model performance is continuously monitored and optimized using Genie and Unity Catalog. According to Bruce, “If we’re building AI into our products, we need a platform that integrates with all our data. Databricks provides that.”
According to Weiwei Hu, TIO’s Director of Big Data Analytics and Insights at HP, this is just the beginning: “Generating real-time, actionable insights has been our primary focus for years, and we see more opportunities Databricks can offer ahead with AI/BI Dashboards, Genie and Mosaic AI.” HP is laying down the foundation for a modern, AI-powered analytics ecosystem — one that can empower teams with self-service analytics and insights to accelerate strategic decision-making and simplify data governance at scale.