Skip to Content
Emerging technologies

Delivering tomorrow’s intelligent process automation, today

Innovation Nation talks to Gina Gray, Commercial Director at Celaton, about how its Intelligent Document Processing platform – inSTREAM™ – helps Capgemini’s customers automate and streamline their end-to-end business processes to deliver a significant reduction in their operational expenditure.

Welcome Gina. Could you start by giving an overview of Celaton and some of the recent changes to the organization?

Gina Grey

Gina Gray: Yes, of course. Over the last few months we’ve seen a number of changes at Celaton, the most notable being the relaunch of our brand, which you can see in all its glory on our website. The new look and messaging reflect not only the increasing maturity of the automation market, but also Celaton’s unique personality, which combines strong technology, expertise, and a passion for supporting customers on their digital transformation journeys.

This new look and approach has enabled us to take the next step on our journey and better represents our people, who are the heart and soul of what we do. This is why we chose to adopt a logo reminiscent of honeycomb – our people, technology, and customers working together in harmony to achieve a common goal, just like bees in a hive. Each element is dependent on each other and vital to achieving success.

It has been an exciting journey since I joined Celaton in 2014, no more so than over the last year. The management team are working closer than ever to ensure that the deployment and development of our Intelligent Document Processing (IDP) platform – inSTREAM™ – is as efficient and effective as possible. This ensures that our customers always receive the best service and technology to support them on their digital transformation journeys while solving real-world business problems.

When I joined Celaton my role was focused on business development by building relationships with partners and any new direct customers. The Capgemini partnership has been our biggest success story with regards to partners and we now have 12 live customers with them and another 4 in the pipeline currently.

My role has expanded significantly and I now have overarching responsibility for the total business revenue including business development, marketing, partnerships, and existing customers. I also work closely with our CTO, Richard Hill, along with the project management, implementation, and support teams to ensure we are able to deliver to, and where possible exceed, customer expectations.

Could you summarize exactly what Celaton does for Capgemini’s clients and the value you bring to our solutions?

When Capgemini initially reached out to us they were looking for an artificial intelligence (AI) solution to extend their automation capabilities in the delivery of services within their BPO division – more specifically invoice processing. Over the four and a half years since the start of our relationship, we have successfully delivered a unique blend of Capgemini’s expertise and Celaton’s inSTREAM IDP platform to Capgemini’s customers.

Organizations are receiving increasing volumes of unstructured and semi-structured content flowing in from customers, suppliers, and employees. With such high volumes of unstructured data, combined with the limitations of legacy business systems, not able to deal with this growing variety of channels, organizations are faced with significant challenges in scaling their operations, enhancing their value proposition, and increasing customer and employee satisfaction. Celaton’s deployment of our inSTREAM IDP platform enables Capgemini’s customers to automate the handling of this unpredictable content through a machine learning (ML) system called “Human-in-the-Loop” resulting in significant resource and cost savings.

Enablement of Human-in-the-Loop means the platform is continually learning from the processing of exceptions, by collaborating with operators who teach it where to find key information. One of the key benefits of inSTREAM versus other automation tools is the exceptions are handled “in-flight” as part of the standard process and do not require separate exception queues or configuration by professional services employees in order for the documents to be processed.

inSTREAM learns from every transaction it processes, and as the volume of documents increases, its confidence and accuracy improve to a point where straight-through processing is achieved with no need for reprograming. As inSTREAM’s confidence and accuracy increases, the amount of interaction needed with humans decreases, enabling Capgemini’s customers to focus on more value-based tasks.

inSTREAM was designed to reduce labor intensive, repetitive, and often error-prone tasks, enabling Capgemini’s customers to spend more time on customer-centric and complex tasks. By doing so, inSTREAM has demonstrated its ability to significantly reduce operational expenditure. On average, Capgemini and Celaton’s Intelligent Process Automation service enables clients to realize 86% reduction in invoice processing time. In addition inSTREAM also enables Celaton’s customers to achieve an average 85% reduction in manual processing of customer correspondence, 75% efficiency gains achieved for sales order processing and a 74% reduction in average transaction costs for claims handling.

How does inSTREAM work?

inSTREAM comprises multiple technologies including AI, ML, optical character recognition (OCR), intelligent character recognition (ICR), natural language processing (NLP), and cognitive computing to deliver the best solution to address a customer’s individual business challenges. Unlike traditional robotic process automation (RPA), which automates tasks within a process, inSTREAM delivers automation to the end-to-end processes such as accounts payable, sales order processing, customer correspondence, and claims.

So how does inSTREAM work? This is easier to explain in the context of a business process such as invoice processing. inSTREAM is able to process all invoices regardless of how they are received including paper (post), email and attachments, fax, and electronic data interchange (EDI). It is able to read and process all formats, structured, semi-structured, and unstructured ensuring all invoices received follow a single process through inSTREAM.

inSTREAM prepares the documents for processing by converting them into a digital, machine-readable format – using OCR if needed – to normalize all documents regardless of its originating format. inSTREAM uses its AI algorithms to apply previous learning and experience to understand the document type, meaning, and intent, and then apply business rules to extract key data. Once inSTREAM has extracted the data it is able to carry out further tasks to validate the data including (but not limited to): verification of supplier details, matching invoice against purchase orders, 2-way matching, date and currency conversion, internal approvals and payment terms, external compliance, checks and measures including duplication, date validation, and invoice values are correct.

Having reached out to the other sources such as external databases or business systems, the verified and enriched data is collated and inSTREAM applies business rules in order to suggest an outcome. Should inSTREAM not feel confident about the outcome or business rules and checks and measures do not add up, the invoice will be shown to an operator (Human-in-the-Loop) to assist with validation and teaching. If, however, confidence and accuracy is high, inSTREAM can deliver the document and output directly to the business system without needing any human interaction (straight-through processing). The outcome could also be the request for additional information or rejection of a non-compliant invoice, which inSTREAM will also perform without needing human effort.

The application of a combination of technologies within inSTREAM to the different tasks within the process, enables the platform to be effective at streamlining end-to-end business processes and deliver transformation benefits to customers.

What’s next for Celaton?

Until now, inSTREAM has been run in a traditional data center in London – when Celaton first developed inSTREAM, clouds were things you found in the sky. But the world has changed, and the cloud has become a proven, stable, and reliable place to operate global services.

With this in mind, inSTREAM is moving to a new home in the Amazon Web Services (AWS) cloud. This move will bring the opportunity to grow our technology and architecture in line with modern trends in a way that is simply not possible in a traditional data center. This provides us with the opportunity to leverage the encryption, security, and reliability offered by AWS to help us continue to deliver our service to the high standards our customers have come to expect.

Our move to AWS has been planned with four key considerations at its core: Security, High Availability, Scalability, and Disaster Recovery. AWS gives us the tools needed to ensure we meet our needs in each of these areas, while also offering the opportunity for future expansion and growth as we explore and take advantage of more of the technologies and global availability that AWS has to offer.

Gina Gray is the Commercial Director at Celaton.