Financial services firms spend countless hours preparing and proofing financial and regulatory statements. Not only are there multiple statements prepared semi-annually and annually, but some larger firms must prepare statements adding up to hundreds of pages. Errors in reports can be embarrassing and even carry penalties. Errors stem from transposition mistakes when compiling reports. Occasionally, statements reveal actual errors in accounting. To address these issues, firms devote considerable internal and external labor to read and reconcile figures and descriptions. We applied several algorithms to read paper statements, convert them to machine readable data, then match and reconcile the information with core accounting systems.
