Penske Drives Excellence and Efficiency with Gen AI Using Snowflake Cortex

Penske is a global leader in supply chain and logistics solutions. Whether they’re coordinating a critical shipment or designing an entire network, businesses rely on Penske’s top-notch services daily for timely deliveries with flawless execution. 

As one of the corporation’s key business units, Penske Logistics offers dedicated contract carriage, warehousing, transportation management and customized supply chain solutions to customers around the globe. With so many constantly moving parts, data is central to Penske Logistics, playing a critical role in optimizing routes, ensuring the seamless flow of goods and reducing disruptions. 

The organization needed a way to manage all its performance indicators in a seamless, scalable way while creating a solid foundation on which to build new AI innovations. Thanks to Snowflake’s robust data platform and the generative AI capabilities through Cortex AI, Penske Logistics has now consolidated all performance indicators while successfully launching generative AI projects that have improved employee productivity and performance.

Penske Logistics is no stranger to driving competitive advantage through data. It was an early adopter of data science to enhance fleet operations and customer service, and the company was among the first to integrate electronic logging and telematics devices into its trucks to better understand driving data and improve safety and service levels. 

To keep track of all this data, Penske logs performance data across different sources. But several years ago, the company found it challenging to effectively use this data due to limitations with its on-premises data solutions, which included an SQL server, and mainframe systems among others. Reporting processes didn’t provide leadership with accessible visibility across the organization, while business analysts relied on other departments to create business intelligence reports. The reporting tool was limited to running in one or two locations using just a month’s worth of data — and would crash if used for extended periods. 

To improve data capabilities, the data science team needed to move to a cloud-based, data lake-driven solution. Snowflake’s ease of use, scalability and built-in data science functionality quickly won the team over compared with competitors. Snowflake's platform-agnostic capabilities allowed the company to overcome the limitations of on-premises systems and achieve a broader view of the organization.

With Snowflake on Azure, Penske Logistics has centralized data in a secure environment, scaled its infrastructure to improve performance, and immediately improved its reporting capabilities. Now, business analysts create BI reports with companywide data spanning five years in just 15 minutes. Additionally, analysts can use certified self-service tools like Tableau to build their own reports. This has freed up the data engineering and analytics teams to focus on expanding service delivery and building out more value-driven features and capabilities.  

The leadership team now uses these reports to gain critical data insights and make more informed decisions. Every department of Penske Logistics makes use of data science through Snowflake, from operations to human resources. Data scientists spend less time accessing data, and Snowflake helps them create repeatable processes for handling and analyzing the data, improving consistency and productivity.

As an added bonus, Ram says, Snowflake has become an incentive for potential prospects looking to join his team: “For data scientists, Snowflake makes our jobs much easier.”

In keeping with its early tech adopter roots, Penske Logistics’ leaders wanted to harness the benefits of generative AI for their operations. To help provide managers with a comprehensive view of performance indicators, the data science team explored various AI solutions that could analyze and consolidate data to look for trends and proactively raise awareness to spark conversation.

Snowflake Cortex AI solved these challenges. Because it hosts and serves large language models (LLMs) within Snowflake’s governed perimeter, Cortex AI made it easy and cost-effective for the team to securely bring LLMs to their data — without requiring additional approvals or investment for new AI vendors.

Safety, performance and retention are paramount for Penske Logistics. Managers aim to improve these areas with personalized insights and recommendations to their teams informed by massive amounts of data. However, insights from this data were not easily accessible and analysis was a manual process to parse through nearly 30 data columns in a dashboard.

Now, Penske uses LLMs in Cortex AI to process performance indicators in record time and produce relative talking points and recommendations for each associate. These insights allow managers to guide conversations with more clarity, precision and actionable next steps.

Since rolling out the solution, managers are enjoying productivity gains from reading a comprehensive paragraph, eliminating multiple sources and columns of data. The company is already seeing improvements in associate performance, safety and retention.  

Penske Logistics plans to integrate Cortex AI’s capabilities into more areas of the business to supplement performance interactions and improve insights. Machine learning models drive several list-oriented processes in the organization, including getting feedback from warehouse associates, and the data science team will use LLMs to enhance sentiment analysis. In addition, the team plans to implement Document AI to automatically extract data from handwritten and printed logbooks, improving operational efficiency and data accuracy.