Starting out on data warehouse projects can be a daunting task, especially in organizations where data demands are constantly expanding. In a video interview recorded at TechTarget's BI Leadership Summit in New York last month, Sam Strum, director of data services at ocean shipping services provider Inttra, offered advice on how to manage that process based on the Parsippany, N.J., company's experiences on a recently completed project to build a next-generation data warehouse and business intelligence (BI) system.
Strum described Inttra as an "Expedia for ocean containerized shipping." Customers looking to transport goods can use its website to check the shipping schedules of various cargo carriers and book spaces for their containers on the voyages that work best for them. Inttra then tracks every step of the shipping process, analyzing data along the way, and facilitates payments between the two parties.
Strum joined Inttra in early 2013, just in time to dive in at the start of the project. "At that time," he said, "I knew that there needed to be changes and there was an opportunity to create a nice data warehouse." The company had an existing Oracle-based warehouse "with a lot of performance problems," he said; a Greenplum data warehouse appliance was also in place, "but it was underutilized." And Inttra had yet to install any BI tools -- instead, "there was a proliferation of spreadmarts," he said.
When planning data warehouse projects, Strum recommends making sure ahead of time that you're clear on what you want to accomplish. "I kind of took a step back and said, 'Let's come up with our principles of what we want to do,'" he said. One of those principles was to retain the Greenplum appliance and better leverage it as the primary data warehouse platform going forward. Strum's team also worked to better unify the company's data by establishing a normalized structure akin to an operational data store. In addition, they took part in the evaluation, selection and deployment of BI software from Tableau Software Inc.
The BI evaluation and selection process also included representatives from the infrastructure management group in IT and a reporting and analytics team that was the business unit involved in the project. The participants chose Tableau's BI tools partly because of the software's ease of installation, according to Strum. Having the Greenplum appliance already installed also helped speed up the project. And, he said, "there was a good alignment between business and IT, so we were able to understand the use cases very quickly."
There's still work to do, though. About 30 Inttra employees are using the Tableau tools now on a semi-regular basis; Strum is looking to increase internal adoption and expand usage externally to both carriers and shippers. More analytical applications are also planned for delivery in 2014. "We're still moving," he said. "We have success, but we're not done."
In the interview, conducted by SearchBusinessAnalytics Executive Editor Scot Petersen, Strum further discussed the BI and data warehouse project and Inttra's technology decision-making process. Viewers of the nine-minute video will:
- Learn more about Inttra's reasons for embarking on the project.
- Hear Strum explain the process involved in choosing and implementing products for the project, including the role of senior executives.
- Get his thoughts on how the company managed to complete the deployment in only 10 months.