"Smart capture technologies allow you to process documents without knowing what you're going to receive."

Smart capture allows the automatic processing of unstructured documents. In an interview with the ICT-Journal, Romain Rochat, Manager Suisse Romande at Arcplace, explains how these solutions work and their most useful applications.

What are the limitations of the previous generation of capture technologies?

For a long time, and this is still the case today for many companies, capture was seen as the simple act of transforming a paper document into a digital image. However, it quickly became clear that an electronic document does not bring much benefit if it is not associated with usable data, thus avoiding the need for manual entry and enabling task automation. To extract this data, ideally automatically, it was often necessary to teach the system where the information was located, sometimes by configuring each document template with the more or less exact position of the data to be read. Although some systems offered more advanced functionality, these solutions remained relatively static and too dependent on prior knowledge of the document to be scanned.

Romain Rochat, Arcplace AG

Romain Rochat, Manager Suisse Romande at Arcplace

What is meant by smart capture? What types of documents are concerned?

Today everything is becoming more "intelligent", from cars to phones, but contrary to the beliefs of some science fiction aficionados, software and robots will not replace humans overnight. Humans must become more intelligent by focusing on the tasks that only they can perform. Smart capture is able to recognize documents and file them, just like employees who sort the mail every morning and put it in trays for distribution to each company department. Sometimes the choice is easy, but often the employee must read and understand the context to ensure correct sorting. This is what smart capture software can offer. It means that all documents can be processed, especially completely unstructured documents. A complaint letter from a customer, for example, whose content and structure are not known in advance.


How do these technologies relate to existing software and data?

The advantage of such a solution is to be able to use existing data from company software (ERP, CRM, customer database, suppliers, etc.) and draw on this information as a major source of knowledge, and thus detect elements on the documents processed without any prior knowledge of what you are looking for. The ultimate goal is to be able to classify documents, i.e. to determine their type, and also to obtain the associated information in order to guide their distribution within the company (thanks to an ECM and workflows system) and ultimately avoid a long and often error-prone entry process.


Can you give some examples of processes that can be automated using these technologies?

One of the best examples is the processing of a company's incoming mail. Some letters are obviously easily recognizable (reply forms, etc.), but most of the documents received are "unstructured", i.e. we do not know what the layout and content will be. With this technology, the first step is to be able to classify documents, i.e. to automatically sort the documents into families according to their content. These documents can be contracts, complaints, invoices, information requests, etc.. Finally, in an incoming mail management solution (also known as a digital mailroom), the aim is to be able to automatically forward mail to the right person in the company, thus replacing the manual work performed by the employee who each morning opens the envelopes and distributes the mail to the various departments, teams and employees within a company. In the case of a bank, for example, a client who sends a letter does not necessarily know the name of their advisor. By using company data, the software will recognize the client (using several pieces of data), detect that it is a request related to their mortgage and, using company data, send the mail directly to their advisor, mentioning that it is a request concerning the mortgage for the chalet in Valais, and not that of the client's main residence.


What types of companies are set to benefit most from smart capture technologies?

It is clear that these technologies, are intended for companies that have a certain volume to be processed. These technologies make sense when automation accelerates manual processing. When the volume of documents received is significant, and a significant amount of time is required to enter, sort and classify them, then this kind of solution is very attractive. We implement these solutions for banks, real estate companies, large industrial and commercial companies and public authorities. Today more than ever, we must do more with less. These smart capture technologies are responses to this paradigm. In the past, capture was sold as a solution for the automation of repetitive and tedious human tasks. This is still valid, of course, but the challenge today is to take it further. These systems can automate processing operations that were previously reserved for human intelligence. In the end, the tasks change, the automation bar is raised, but the human mind, our creativity and irrational thinking are still required to perform tasks and make decisions which cannot be completed using software or an algorithm.

 More about "Smart Capture"