Hybrid architecture models
E-3: The pandemic has given digitalization a huge boost, can you confirm this?
Matthias LemenkühlerIn any case. In many companies, important business processes came to a standstill in spring 2020 with the abrupt switch of a large proportion of the workforce to working from home - precisely because they were still in paper form and a lot of work was done manually.
If an invoice is in a document folder on the desk in the office, it simply cannot be checked and approved there. In no time at all, and with many workarounds, companies have tried to digitize these processes if they were not already planning to do so.
E-3: Will it stay that way?
LemenkühlerDigitalization and the cloud have gained enormous acceptance as a result of the pandemic. Even if the situation has eased, I don't think the clocks will turn back completely. The benefits of digitalization have now become all too clear.
E-3: Should companiesn now fully rely on the cloud?
LemenkühlerAnyone who really wants to work digitally from end to end, including when exchanging information with business partners, can only do so by using the cloud. However, the status quo is that the cloud is not yet the predominant standard. With our customers, we actually always find the situation where there are existing processes and systems installed in-house, which is also referred to as "on-premises". Especially in the area of business software, i.e. ERP systems, there are structures that have grown over the years and do not need to be radically replaced.
E-3: What do you recommend?
Lemenkühler: Hybrid operation is recommended, i.e. mixed models of on- and off-grid operation.
premises and the cloud. This is where we come in and help companies to make their document-based business processes simpler, more flexible and faster - by adding certain services from the cloud. Our strategy is: "Cloud first, but not cloud only".
E-3: Does using an ERP system automatically mean paperlessit work?
Lemenkühler: It would be nice. Of course you want to carry out work processes from procurement to payment within your ERP system. However, this does not mean that they are automatically completely paperless. Ideally, workflow and process solutions such as those from the xSuite Group take care of digitization. They also ensure that document-based processes are mapped automatically throughout the ERP system.
E-3: Thank you for the interview.
What's the hype about AI and machine learning?
Invoices have been written, checked and paid ever since people have been doing business with each other. A traditional procedure that is nevertheless subject to change - in the recent past mainly due to digital technologies.
Digitization in invoice verification sounds progressive. However, many people in Germany still only understand it to mean scanning incoming paper invoices and processing them electronically. Many countries around the world are already further ahead. Invoices no longer need to be scanned because they already arrive in digital form, by email or PDF.
In the meantime, however, the trend away from paper and towards e-invoicing is unmistakable in this country too. Those who work digitally from the outset can focus more on the actual automation of processes.
On the one hand, this means digital workflows for checking and approving invoices.
At the same time, artificial intelligence (AI) technologies are supporting the further expansion of automation. AI is not an ominous hype topic here, but can be used in concrete terms to relieve employees from purchasing and accounting of routine tasks.
An invoice solution, for example, not only reads the contents of an invoice, but also "understands" them and can make suggestions as to who the vendor is, which account assignment is suitable, etc. You then only have to choose from these alternatives without having to memorize or look up all the account assignments.
In order to make good suggestions, AI applications need to learn. Machine learning works all the better the more data records are used. This is why AI technologies are often used in public cloud solutions. These are usually designed for a large number of clients. Even if the individual clients' data is logically and physically separated, the system still has the option of using all data records as a learning set.
AI can also be used for analysis: Which are the most reliable suppliers or cheapest prices, when is the best time to order? AI can derive all of this from invoice documents and provide valuable tips. Artificial intelligence can also be used to automatically detect attempted fraud: Where are there a conspicuous number of invoices or those with above-average amounts? AI takes processes to a whole new level, not only in terms of automation, but also in terms of analysis.
DE
EN
ES
