Mining Actionable Information with Smart Capture: INTERVIEW WITH ENNO LUECKEL, EPHESOFT

Companies collect tons of data nowadays, but most of them do not yet profit from that treasure. The situation can be compared to the peak of the California Gold Rush: all of the easily accessible gold had been recovered already and mining the rest had become very expensive, but then new technologies were invented to recover the metal cost-effectively. This is more or less exactly what Ephesoft did for data: while the cost to recover actionable information has been very high in the past, Ephesoft introduced machine learning tech­nology for data capture, which makes extracting data more affordable and easier to access for organisations of all sizes. We talked about this approach with Enno Lueckel, Vice President EMEA at Ephesoft, discussing how it works in daily business and what it means for companies.


There are several data capture suppliers on the international market. What is the difference that makes Ephesoft’s solutions unique?

We disrupted the document capture industry by moving from simple data collection to intelligence. Most other suppliers focus on scanning documents, while we introduced Smart Capture to gain actionable information for our customers. Our solution is the first platform which can be run on-premise and as a cloud-based Capture-as-a-Service platform, which means lower entry costs, flexible architecture and easy imple­mentation for organisations of all sizes. Additionally, we combine classic methods of extraction with super­vised machine learning. Initially, this approach uses human trainers to validate the classifications for a few days, but every input makes the system better, faster and more accurate. For example, a recent client engaged in supervised learning for only two days, and we delivered a 92 percent increase in accuracy and 400 percent increase in productivity over what they had achieved with a competing service before.


Due to the approach of combining hybrid cloud and subscription model, our solution is not only affordable for big enterprises but also for small- and mid-sized companies

That sounds quite interesting! Which indus­tries profit from that solution?

Any organisation that processes a wide range and large number of documents can profit from our solutions. In some companies thousands of docu­ments are processed manually each day, week or month. This includes – for example – companies in the financial services, banking, insurance, healthcare, accounting and invoicing industries as well as govern­ments and the education sector. Due to the approach of combining hybrid cloud and subscription model, our solution is not only affordable for big enterprises but also for small- and mid-sized companies. Also, our solution allows organisations, which currently store their unstructured data in dispersed storage centres, to merge this hard-to-access data into action­able information. This improves accuracy as well as transparency and reduces risks.


This transparency is also impor­tant in light of the General Data Protection Regulation (GDPR) being rolled out by the European Union. How does Ephesoft help with that challenge?

That’s right. Facing the GDPR, every company must be able to identify every piece of a customer’s personal data. This includes information from paper files, emails, PDFs and all other channels, including backup files. This unstructured data within different documents and various platforms constitutes a high risk and can easily turn from an untapped asset into a major liability. Companies that do not comply with the GDPR are in danger of hefty fines. To prevent that, our solution Ephesoft Transact ensures the necessary transparency and inte­grates document capture functionalities into ERP, CRM, and ECM systems as well as back office programs. Examples include SharePoint, Office 365, Infor, SAP, Box, Alfresco, Salesforce and many others.


You have announced the new version Ephesoft Transact 4.5 – what are the main changes coming with this launch?

The biggest news is the many features, which are using more and more super­vised machine learning (SML), starting with Web Services, which provide the flexibility to integrate SML into existing document capture solutions to gather metadata fields. Additionally, the tech­nology will provide enhanced extraction of tables and line items. This reduces the amount of configuration that an admin will need to set up in advance. Operators can just push a button in Ephesoft Transact and force the system to “learn” the table in question. When it sees this type of table in the future, it will know which values to extract and how to extract them more accurately. We also improved the processing speed: at the most basic level, the Ephesoft platform runs on a 4-core processor and customers can upgrade to an 8-core, 16-core or higher, depending on the necessary perfor­mance speed and number of documents they are processing. Naturally, customers with higher-end hardware will see a more meaningful performance increase: a 10% increase in PPM (pages per minute) on an 8-core and a 32% increase on a 16-core PC, while high-end hardware can get you even higher figures. For example, a system with three 16-core CPUs can get a whopping 64% performance increase with Ephesoft Transact 4.5. In this scenario, customers can process almost 550 PPM. That’s significant. Of course, this number only increases as you add more cores and servers (we tested three servers with a 32-core CPU and it yielded almost 750 PPM).


Our mission is to enable organisations to make better decisions by mining the world’s unstructured data and transform them into actionable information.

Will there also be new developments in regard to third-party-programs?

Yes, we introduced a new Technology Partner export plug-in functionality, which increases flexibility and is especially interesting for medium-sized and large enterprises. We integrated Box, which is becoming quite a popular repository for documents. Users can use Ephesoft Transact to capture, classify, and extract metadata from their documents, then send that information directly into a Box repository. Organisations won’t have to write any custom code to set this up. Additionally, with SharePoint we integrated one of the document repository services with the largest share of the market. This also func­tions as an out-of-the-box connector, including comparable features. For ERP, we partner with Infor, but also integrate into SAP, providing different scenarios for linking a document to a specific BusinessObject instance, and SAP BDS. On a project level we also offer integrations with SAP DMS and SAP Folders Management.


Those sound like great opportu­nities! So, what will the future bring for Ephesoft’s customers?

Our mission is to enable organisa­tions to make better decisions by mining the world’s unstructured data and trans­form them into actionable information. Therefore, we focus on our customers, we try to innovate new products every two years, and we always use the latest technology available.


About the Interviewee

Enno Lueckel leads Ephesoft’s Sales teams in EMEA. He brings over 15 years’ experience in the areas of document capture and enterprise content manage­ment. Prior to joining Ephesoft, Enno built Notable Solutions’ European opera­tions, transforming the organisation from a technical sales approach in the MFP industry, to an EMEA-wide provider of decentralised capture solutions.


Please enter your comment!
Please enter your name here