Supply chain and cargo operator LINX Cargo Care Group has created a cloudy data lake and given line-of-business people who own different apps the keys to the new tool.
LINX is a new company formed after breaking away from the ASX-listed Asciano group of companies. As such it was able to retain some legacy applications and data, but quickly adopted a cloud-first strategy for new infrastructure that has seen PeopleSoft replaced by Oracle Cloud and other legacy software replaced with, or moved to, SaaS.
To make the move work the company migrated legacy data to new cloud-native apps, after ingestion projects made tricky by differing data structures.
After a few ingestions LINX General Manager for IT Thom Gianniodis felt the company had enough data to create a data lake, plus a business case that made it financially viable.
He’s since created a group of “data insiders”, line-of-business execs from across the business who own different applications. The insiders direct the flow of data into the lake and are also being trained to use it without IT's help.
Gianniodis has appointed five insiders, one each for key business applications, and convenes a monthly meeting at which he and the IT team “explain the data catalog and how it works.”
“As good as our IT people are, they do not get the operational business skill,” Gianniodis said. Sessions with both groups present mean that line-of-business people can tell IT what kind of data will improve the lake’s utility. In one instance, business owners explained that fuel price fluctuations would help them to make better decisions. Gianniodis’ IT team was able to source that data, add it to the lake and improve decision-making.
The insiders are also being trained in Microsoft Power BI and the Visualiser service that operates with Azure Databricks and Azure Data Lake Storage. The training means the champions can do self-service analysis, which makes the lake more valuable again.
The tool’s already produced some wins: Gianniodis said analysis of current and historical invoices lets the company understand its creditors.
“Instead of reporting out of Oracle we are bringing that data into the Lake and made a dashboard to see late payers,” he said. “Then we went to predicting who would not pay the next month.”
“We have essentially gotten that to 95 percent to 96 percent accuracy.”
The company uses those predictions to inform customer relationship managers so they can talk to clients about business issues, rather than conversations purely about paying bills on time.
“We are getting a little more proactive,” Gianniodis said.
The lake is also helping the company to understand the impact of events that mean it has to re-schedule pickups and deliveries, or how weather at sea changes shipment times for the bulk goods – think grain, metals or and cars – in which LINX specialises. Analysis of those changes means better-informed customers and, again, more pleasant conversations for relationship managers. The lake has also revealed secrets of staff leave liabilities and progress towards budgets.
The company’s choice of Azure for the new facility came from an odd direction: Asciano was a Microsoft shop and a user of hosted Citrix desktops. LINX persisted with the latter, hosted in Azure, and enjoyed the experience enough that considering Microsoft’s cloud for other duties was a a natural step.
That probe led Gianniodis to believe that Azure “was more complete and it did not need as many developers. With AWS we would have needed more developers than we would have liked.”