Molson Coors Implements Cognitive Invoice Data Capture Tool
Molson Coors is a multinational brewing company whose origins date back over 230 years. Famous brands span the globe with products ranging from beer and cider, to non-alcoholic beverages and coolers. Ranked #5 for the world’s largest beer companies, and #5 for worldwide sales in 2018. Molson Coors has shaped the brewing industry and has unparalleled staying power.
What was happening with Molson Coors?
Molson Coors was looking to automate a function of their business processes throughout the whole group of companies. Invoice processing was the most urgent issue, as the manual work was very tedious and time-consuming, and there was the most space for automation.
Who did Molson Coors choose to automate invoice processing?
EY, acting as an RPA integrator, decided to implement Blue Prism robot to help in Molson Coors’ situation. EY offered Molson Coors two options for invoice processing – a traditional OCR provider or Rossum. Molson Coors chose Rossum, the invoice data capture tool, which can work without any templates thanks to Rossum’s artificial intelligence technology.
How Molson Coors benefited from Rossum app:
After an in-depth analysis, Molson Coors decided to proceed with Rossum – particularly due to the following reasons:
- No templates – Rossum does not require the setup of any rules or templates, which decreases costs of deployment as well as continued maintenance costs.
- Fast deployment – From deal confirmation, to deployment of the app in the Molson Coors infrastructure, the process took less than 3 days. This also included connecting Rossum and Blue Prism robot, setup of fields to be extracted, and provisioning of 3 Molson Coors accounts to the cloud-based user interface.
How Rossum automated invoice processing for Molson Coors:
Blue Prism robot begins by gathering invoices and sending them via email, where Rossum’s AI engine would receive them and capture the invoice data.
- If there was high confidence data, it would be passed back to Blue Prism robot where the extracted data would be matched with the POs. The robot would then store the results in the targeted system.
- If there was low confidence data, it would be sent for validation with Rossum’s user interface, which would be presented to the human operator. After, the app would update the captured data and send the validated results to the targeted system.
Molson Coors started their internal testing and is now using Rossum as a backend function to input structured data from invoices into SAP. Since implementing Rossum, it is possible to extract invoice data fields without the tedious manual data entry that was once necessary. Customized Rossum app was deployed into Molson Coors’ infrastructure in only 3 days, and extraction accuracy has reached 97.3%. Invoice automation is now possible throughout the Molson Coors ecosystem, thanks to Rossum’s AI.