In order to remain competitive in the crowded arena of financial lending, Queensland-based firm Suncorp realised it needed a strategy to accelerate loan processing.
The company has 16,000 employees and about 9 million customers across all aspects of the business, which includes banking, investments, superannuation and insurance.
Suncorp executive manager enterprise applications, Dawn Stephenson, said when she came into the role in August 2010 there was no solid strategy around content capture of data, while there was a growing demand from multiple business units for the digitisation of documents.
The company had recently invested in IBM’s FileNet platform for enterprise content management and it was also using it for business process management (BPM).
The search for new content capture software was rigorous, Stephenson said. Suncorp shortlisted three candidates: EMC’s Captiva, IBM’s Datacap and Kofax.
“The [Kofax] product ticked all the boxes, but one of the main reasons we went with it was that I could see there was a lot of investment in R&D in the product itself; whereas I couldn’t see where that same investment would be made with the other suppliers,” she said.
“There was also a strong professional services arm here to support us through the implementation which the others didn’t necessarily have.”
The goal, she said, was to reduce the use of paper, the storage of paper and the cost of the real estate needed to store the paper, while increasing compliance around document storage.
The first cab off the rank was Suncorp Bank’s loan origination department, as it had “a burning need” to automate its end to end loans process.
“They were working with a lot of paper files so at any one time there could be5000 paper files around the building,” she said. “Paperwork would come in, get printed off whether by fax or email then put into a folder which would move through the business processes from operator to operator.”
“The amount of paper that was being used and the tracking of those files was a challenge to say the least.”
In addition to eliminating paper, the bank also has a requirement to locate and view documents quickly, driven by the concept of “showstoppers” — referring to critical information for a loan which would stop the application from progressing.
“Kofax can recognise that our loan application form is 15 pages long… It captures the document in a TIFF or PDF file and then it can classify what that document is and then extract information from there,” she said.
It can also quickly register if mistakes have been made in the documentation and quickly rectify them; previously mistakes were not identified until after printing had taken place. Some forms are handwritten, but the software also has character recognition capabilities.
Implementation and testing took two months, with the paperless phase of the end-to-end automation of the loans process in place by July this year and the storing of documents in FileNet still in the pilot phase but due for completion in six weeks. “It has enabled the team to get rid of all the printers they had on the floor, and eliminate the 5500 files they had at any one time which required several people to carry them from team to team within the process and the storage costs associated with the files are now nonexistent.”
The plan behind the company’s $1.5 million investment in the Kofax platform was to roll it out across all Suncorp business lines that have a need to capture data. This encompasses the human resources, insurance and finance units as well as potential demand driven by a significant upcoming banking transformation project which will take some years to complete.
“Implementing this has enabled us to capture and store content from end to end and to respond to our customers quicker in the loans department. We can turn around information a lot quicker so we don’t lose out to other banks due to long processing times."
According to Stephenson, special attention should be paid to form design as it could result in the software being unable to classify the information automatically. This resulted in quite a bit of form redesign, Stephenson said.
Of the documents going through the system, around 50 per cent are unable to be classified automatically at this stage, Stephenson said, as the bank continues to err on the side of caution.
“You set the confidence levels around the classification, anything that the software isn’t sure about is output to an exceptions queue and then a person will manually classify that document.
“A lot of documents go to the exceptions queue but that’s part of the form redesign and learning about what those forms are.”
As an [artnid: 361407|avid Agile supporter|new]], Stephenson said the techniques helped get the implementation across the line in the two month period, despite the Kofax team being inexperienced with the methodology.
“Something I would make sure of next time is that they had some Agile training because we really had to walk them through how we wanted to implement it and look at whether they could work with us in that space.
“With hindsight I would have made sure that Kofax beforehand understood that we would be working Agile and what they might need to think about in how they were operating as well.”
Follow Chloe Herrick on Twitter: @chloe_CW
Follow Computerworld Australia on Twitter: @ComputerworldAU