Responsible companies are not only forthcoming with the data they store and use, but also meticulous about keeping their data tidy and accessible. Because messy data can lead to non-compliance even at companies run by people for whom fabricating data or concealing information is completely off the table. But in today’s world of rampant data overload, even the best-intentioned among us sometimes get inundated with paperwork. Figures wind up wrong, data winds up hidden. Luckily Applica utilizes an important data automation feature that can help not just with the speed at which your company can process the paperwork, but also with the way this paperwork is kept tidy, accurate, and compliant with the rules.
Our innovative Business Context Identification (BCI) capability was originally designed to assist the assignment of documents to existing accounts. From financial institutions and insurers to telecoms and energy companies, our clients live and die by their ability to assign the right documents to the right accounts. Understanding the need for clients to stay ahead of some common errors in electronic and digitized documents, we had to pre-empt the file identification and case number assignment errors that often result from simple writing and typing mistakes or the vagaries of digitization. We needed our AI to retrieve the correct account reference numbers and assign documents to the right accounts on file.
Unlike our competitors, we were able to make headway here thanks to the way our AI is able to concurrently process structured text, unstructured text, and graphic features, such as signatures and layout cues. We found a way to prevent incorrect document assignment even in the face of some misleading info, especially of the type entered by hand, either in messy handwriting or as a result of hasty typing. The idea was to teach the AI to essentially “check its work,” so that it could be vigilant about all those ones that look like sevens (and vice-versa) and some inevitable—and thus common and predictable—typos, to name just two crucial examples of where things typically go wrong. In addition to preventing outright errors of document mis-assignment, our goal has been to reduce the number of documents our AI couldn’t process, which had previously been deferred to human workers (ones just as prone to mistaking a one for a seven as the AI).
The way we designed the solution harnesses the redundancy that naturally exists in most documents. Significantly, this includes not just obvious cases of redundancy, where snippets of data repeat across the page or on different pages, either because the same data recurs in multiple areas or because different entries code the same data in partial ways (think date-based ordinals in file naming). Such examples obviously allow for multi-step review. But our solution goes further, also allowing such cross-verification of data as is possible through review of co-existing data. Though there may be no explicit informational overlap between certain data points, their co-occurrence may itself repeat across documents. And this is another way that our AI is able to correctly identify business documents using all the contextual cues available.
So what happens when our ID-assigning feature is deployed? Our clients get the intended result as documents are assigned the right reference numbers with significantly more accuracy than would be possible without BCI.
But there’s more. In the course of cleaning up the document assignment workflow, Applica enables an additional layer of benefits: all of a sudden, minor errors can be fixed at an early stage, outdated information can be updated, future hassle or liability can be nipped in the bud. The bonus benefit of our AI’s new feature is as important and useful as its primary benefit, with gains in security and compliance matching the gains in speed and accuracy of document processing.
By cross-referencing the data it “reads” as it processes it, our technology audits your data for you concurrently with processing its source document. And this is typically information you didn’t even know needed auditing, because you had no reason to suspect that any of it is incorrect. Why is this so significant? Because even if you go out of your way to stay compliant, it’s easy to miss the kinds of errors our BCI is trained to see. Though most are innocent mistakes—they can still create delays, precipitate discontinuities, and result in legal trouble. Wouldn’t it be much better to get the right document automation partner to take care of all these problems before they even happen?
For swift document automation, game-changing scalability, and help auditing the things you didn’t know needed auditing—make Applica part of your toolkit today.
