DS Smith produces packaging solutions for companies around the globe. To do so, it spends £3.8 billion annually on procurement. Legacy systems, multiple locations, manual processes, and poorly categorized data were making it difficult for the company to analyze its spending and making it difficult to optimize costs. The company also found it difficult to use procurement data to drive strategic business decisions. Analytics employees were spending up to 60 percent of their time on manual tasks, such as mapping spends to categories.

“Our procurement analytics team was conducting a significant amount of work, chasing people to get the data needed,” says Jackie Ross, IT director for DS Smith. “The data then had to be cleansed and consolidated, which could take a substantial time to achieve reports which ultimately reviewed what had happened. We needed to find a way to look forward.” Procurement was enthusiastic about using technology to transform the way it worked, so Jackie Ross tasked Adrian Pinder, head of digital and data at DS Smith, to source a supplier that could help it gain fast, integrated access to its data and produce timely analytics. “It couldn't be more of the same,” says Adrian Pinder. “We needed technology, and particularly digital and data technology, to be the next growth platform for the company or we’d miss opportunities for growth. We knew we needed to bring in an external partner with that build capability.”

Mathieu Finck, head of procurement operations, South Europe, for DS Smith saw the daily struggles that having multiple, unconsolidated ERP systems caused for his team. “Simple things like knowing how much you are buying from a volume or cost point of view from a supplier was time consuming and imperfect,” says Finck. “We’d have to ask local procurement teams to get volumes or prices. Sometimes, we’d go directly to our suppliers to get the data. I welcomed the idea of consolidating data and improving access.”

Adrian Pinder found Exponentia.ai based on a referral and chose to work with the AWS Partner after meetings and discussions. “They really felt like the right fit,” says Adrian Pinder. “They had a flexible attitude and it was clear that understanding our business and getting the right solution for us mattered to Exponentia.ai. They had the deep data governance and data engineering expertise we needed.”

The initial ask of Exponentia.ai was limited, focused on helping the company produce quarterly reports for the procurement leadership team, but as Exponentia.ai looked at DS Smith’s data landscape, the scope grew. “They didn’t just have issues with getting the data—there were problems with the data as well,” says Sanjay Kumar, chief data and AI operations officer at Exponentia.ai. “There were multiple sources of data, but it was often miscategorized, with different locations using different categories. We agreed to integrate the systems for access, clean the data, and fix the categories. Doing that let us build a true picture of spending that could be parsed as needed.”

DS Smith and Exponentia.ai chose to build the PSA solution on AWS because the company was already using AWS in other parts of the business. The solution uses AWS Glue—a serverless data integration service that makes it easier to discover, prepare, move, and integrate data from multiple sources for analytics, machine learning (ML), and application development—for the extract, transform, and load (ETL) layer, to collect data from 12 source systems (primarily SAP) in the company. AWS Lambda—a compute service that runs code in response to events and automatically manages the compute resources—is used to move the data.

Data is stored using Amazon Simple Storage Service (Amazon S3), an object storage service that offers industry-leading scalability, data availability, security, and performance. The solution uses Amazon Redshift, which can analyze structured and semi-structured data across data warehouses, operational databases, and data lakes, for its analytics. “We learned a lot building this with the knowledge transfer that came from building this,” says Adrian Pinder. “Running the system is a shared responsibility and we have Exponentia.ai people ready to slide into our team and work together whenever we need them.”

Since the first phase of the project went live at the start of 2024, DS Smith has seen a host of benefits. Procurement is able to get near real-time access to data in a single dashboard that pulls from the 12 different sources, and it can do so knowing that the data is properly categorized, ensuring that nothing is missed. By using templates and automated workflows, the company has seen a 15 percent increase in percentage of properly mapped procurement spends.

These insights can have a significant payoff. DS Smith has found that it can realize cost optimization of up to 1 percent on its procurement spending because the new insights it has gained enable it to negotiate from a stronger position. For example, one category manager found a supplier that was being paid $10 million dollars per year and was able to renegotiate to cut $3 million in costs going forward. Another category manager found a telecommunications supplier that had 60-day payment terms but was submitting invoices after only 30 days. Enforcing the terms of the agreement freed up £300,000 in working capital for DS Smith. “Using AWS is fundamental to driving our digital transformation and operational excellence,” says Scott Civzelis, chief information officer at DS Smith. “Like our procurement spend analytics solution, we are using AWS technologies to enhance sustainability, agility, and innovation across our global supply chain, ensuring we deliver unparalleled value to our customers."

Phase two of PSA, which will add more data sources, is under way and DS Smith and Exponentia.ai are also looking at ways in which they can use AI to improve analysis. “It’s remarkable what we’ve accomplished working with Exponentia.ai on AWS,” says Adrian Pinder. “We’ve received great results and great value. The PSA is a hit with the procurement team, but it’s also excited others in the company. Business units are approaching me asking when they’re going to get their own transformation.”