Game-Changer: How Databricks Supercharged the Pacers’ AI Strategy, Cutting ML Costs by an Astounding 12,000X and Accelerating Insights!

Game-Changer: How Databricks Supercharged the Pacers’ AI Strategy, Cutting ML Costs by an Astounding 12,000X and Accelerating Insights!

Transforming Fan Experience: How Pacers Sports‍ and Entertainment Utilizes ​Data Analytics

In the realm of basketball, statistics⁤ are crucial — but for⁣ Pacers‍ Sports and Entertainment (PS&E), understanding ⁤fan‌ data is equally significant.

The Shift to Enhanced Data Insights

The parent ‍organization of the⁣ Indianapolis Pacers (NBA), Indiana Fever (WNBA), and Indiana‌ Mad Ants‍ (NBA G ​League) previously invested a substantial $100,000 annually into⁤ a machine learning platform aimed at developing predictive models regarding ticket demand and pricing. However, the results‍ were not arriving quickly enough.

This scenario propelled Jared ⁣Chavez, the‍ data engineering manager,‍ towards adopting Databricks on Salesforce approximately 18 months ago. The shift has yielded remarkable results; his team⁢ now executes ​extensive predictive analyses with meticulous compute configurations for only $8 per year—a staggering reduction that he attributes to optimizing their⁤ machine learning processes.

“We excel​ in fine-tuning our computing resources to discover just how much ‍we can diminish ‌expenses while still effectively running our models,” Chavez ‍shared with VentureBeat. ‍“This is what‍ defines our success with Databricks.”

Cuts in Operational Expenditures by Nearly 98%

Beyond its three basketball franchises, PS&E also manages an esports venture under Pacers Gaming and organizes numerous events throughout the⁤ year at Gainbridge Fieldhouse—hosting everything from March Madness games⁢ to concerts. Recently, they unveiled​ plans for ‍an impressive ‌$78 million Indiana Fever Sports Performance Center scheduled to open in 2027, connected by a skybridge to both the ⁣arena and parking facilities.

This growth generates‌ massive‌ amounts of data—and challenges related to data management.​ Chavez‍ highlighted⁢ that until two years ago, PS&E relied on two ‌separate warehouses using Microsoft⁤ Azure Synapse Analytics. This led different teams within the organization to apply various analytics ⁣methods with considerable inconsistency across tools and skill sets.

While Azure Synapse could effectively connect external‍ platforms, it was financially untenable for a company like PS&E. Moreover, ⁢integrating ​their machine learning‍ platform ⁣resulted in fragmentation within Microsoft⁣ Azure Data Studio.

A New Direction in Machine Learning

To overcome these challenges, Chavez transitioned operations over to Databricks‌ AutoML along with Databricks Machine Learning Workspace in⁤ August 2023 focusing initially on configuring models‍ pertinent to ticket prices ⁤along​ with​ game attendance forecasts.

This new setup proved beneficial not ‍just for technical staff but also non-technical⁣ users expediting ML workflows while drastically⁣ cutting costs. “‘The speed boost improves response times dramatically ‍for marketing staff who don’t need coding skills,’” stated ⁣Chavez about leveraging intuitive user ⁤interfaces where all functions consolidate seamlessly ​into unified ⁢records through Databricks.”

Data Management Transformation

The company’s extensive array ​of systems was brought under Salesforce Data⁣ Cloud’s ⁢umbrella; consequently enabling them access over 440⁢ times more stored data alongside eightfold increases‍ in operational data sources.

Pioneering ⁣this approach has driven operational expenditures down ⁣below⁣ 2% of prior figures—Chavez noted ‌savings amounting to hundreds of thousands annually which have been redirected towards enhancing customer ‌insights as well as refining tools across analytics departments within PS&E.’

.

A Culture​ Shaped by Continuous Improvement

“How ⁣did we achieve such remarkable reductions?” questioned Chávez⁣ rhetorically before explaining how adjustments made over time ⁣involving cluster settings improved connection capabilities further⁢ streamlined model‍ integration back into centralized databases.”​ He⁣ emphasized that through this ongoing refinement ‍process powered by robust ML engines keeping⁣ every customer interaction up-to-date could⁢ drive higher accuracy earlier each cycle.

Exit mobile version