DDespite going paperless in 2000, Sovereign Bank's consumer lending document imaging and indexing remained a labor-intensive process, says Alex Norris, the Wyomissing, Pa.-based bank's assistant VP. "By late 2006 we had 20 employees dedicated just to imaging," he relates. "It was time to automate."
Sovereign ($77.3 billion in total assets) initially explored imaging solutions that offered an indexing module, including its existing vendor. But the up-front investment was steep, and ongoing volume-based monthly charges would have offset any savings, Norris reports. So in December 2006 Norris floated an alternate approach.
"We had previously purchased ExpressRecognition Server by Adlib Software for a short-term imaging project," he notes. "We knew Adlib lacked the indexing piece, but we were otherwise familiar with the product's capabilities. And we already owned a copy. So I suggested we convert to Adlib for imaging and develop an indexing module in-house." According to the Burlington, Ontario-based vendor, ExpressRecognition provides server-based optical character recognition to support high-volume document workflows.
Once testing proved encouraging, Sovereign, a subsidiary of Banco Santander (US$1.33 trillion in assets), dedicated a developer to writing the code. "We programmed in [Redmond, Wash.-based Microsoft's] .NET," says Norris. And, "We purchased EMC's [Hopkinton, Mass.] PixTools to enhance scanning and processing capabilities."
The project's only real challenge originated with Sovereign's implementation strategy. "For imaging speed and storage efficiency, the project team elected to scan all documents as black-and-white," Norris recounts. "Later we were directed to maintain the integrity of driver's license photos by rendering the images in gray scale. To accomplish this, we developed a workflow that scanned the driver's licenses separately and, using Adlib, automatically matched up the licenses with the appropriate loan account."
By May 2008 Sovereign determined it would need two more ExpressRecognition site licenses. "Because Adlib is lightweight, it loaded onto existing [Microsoft] Windows-enabled servers," observes Norris. "But it's CPU-intensive, so we needed three servers to provide the necessary throughput."
After a month of testing and a smooth go-live in August, the new system quickly proved transformative, Norris says. "Imaging and indexing turnaround fell from over two days to the same day or, at most, overnight," he relates. "Among other things, we've reduced paper costs and scanner time. ... We've reduced imaging head count from 20 to 16. ... [And] now we're indexing an average of 12 pages for every loan instead of only the three most critical pages."
If ExpressRecognition has a weakness, according to Norris, it's the solution's generic error messaging. "It indicates a document failed to scan, but not the cause," he explains.
Today Adlib runs unattended 24/7 on four servers and processes 95 percent of consumer lending paperwork automatically, Norris reports. "For the remainder, the Windows-based interface simply presents the unrecognized image to an operator who types in the outstanding information," he says.
"[ExpressRecognition] hasn't required any IT time beyond product upgrades," Norris continues. "And we can scale it to whatever size we need. So we keep finding new uses for it, such as turning thousands of pages of paper reports into searchable PDFs. ... Not surprisingly our success has attracted the attention of other departments."
Case Study SnapshotInstitution:
Sovereign Bank (Wyomissing, Pa.; a subsidiary of Santander, Spain-based Banco Santander).
Automate document imaging tasks to streamline processes and achieve efficiencies.
Adlib Software's (Burlington, Ontario) ExpressRecognition Server imaging solution.
Anne Rawland Gabriel is a technology writer and marketing communications consultant based in the Minneapolis/St. Paul metro area. Among other projects, she's a regular contributor to UBM Tech's Bank Systems & Technology, Insurance & Technology and Wall Street & Technology ... View Full Bio