As one of the world’s foremost providers of consultancy, technology and outsourcing services, Capgemini constantly strive to deliver enhanced business value, increased efficiencies and an exceptional customer experience to their clients. One of Capgemini’s clients was looking to transform the way they process their customer invoices and requested an innovative solution. Capgemini responded by engaging their Intelligent Process Automation partner Celaton to deploy their Intelligent Document Processing (IDP) platform, inSTREAM™.
67%
reduction in the number of FTEs needed in the process
"Capgemini and Celaton can clearly demonstrate tangible, innovative and fully operational AI implementations that are delivering business value."
Delivering Digitial Transformation with Intelligent Process Automation (IPA)
Capgemini Achieved:
Error reduction - inaccurate data prevented from entering the finance system
Faster processing - ensuring SLA's are met
Significant improvement in customer experience
Competitive advantage through innovative technology
Intelligent Invoice Processing
Delivered as a service, inSTREAM applies its IPA technologies, including Artificial Intelligence (AI), Machine Learning (ML), Robotic Process Automation (RPA), Optical Character Recognition (OCR) / Intelligent Character Recognition (ICR), Natural Language Processing (NLP) and cognitive computing to streamline the processing of invoices for Capgemini’s client.
Upon receiving an invoice from a dedicated Capgemini mailbox, inSTREAM is able to learn the pattern of unstructured content through the natural consequence of processing the document. As a result, inSTREAM is able to read and categorise the document whilst extracting key data and carrying out further tasks such as validation and verification of supplier details, PO checks, currency and date conversion and other business checks.
As a result of applying inSTREAM to their client’s invoice processing, Capgemini achieved a 67% reduction in the number of FTEs needed in the process. The amount of incorrect data generated by human error was significantly reduced and as a result, there was a notable increase in performance accuracy.