Orchestrated workflows and robots as the key to successful automation of insurance processes

Digitalisation, automation and artificial intelligence (AI). Rarely has the technological landscape offered so many changes and innovations as it does today. Only those who use these technologies efficiently and effectively will remain competitive in the long term. Poor integration, unclear, complex processes and incompatible legacy systems pose massive challenges for insurance companies. Insurmountable?

In my blog post, I will show how workflow automation can be used to tackle these problems and achieve a fully automated process.

Digital, but also automated?

Which insurer can claim to have fully automated its own digital business processes? True end-to-end automation, without human intervention, without process interruptions where information has to be added later? For most insurers, this is still a long way off. Because although the digital world is crammed full of technologies, ideas and automation, the reality is often very different. Fully automated processes are not the norm, but the exception.

The problem of untapped technological potential

Insurance companies that rely on end-to-end automation have recognised the decisive competitive advantage. On the one hand, customers no longer want to submit their claims on paper or wait weeks for their claims to be processed while being kept in the dark about the status of their claims. On the other hand, insurers cannot afford to purchase expensive technologies that cannot be used optimally due to a lack of infrastructure or inadequate business processes.

But what exactly is the problem? With mobile apps, multichannel approaches, document understanding, RPA and workflow automation – to name just a few technologies – a fully automated world should have been a reality long ago. However, the theory often fails in practice and can be broadly traced back to three problems:

• Lack of integration of existing technologies
• Opaque, rigid, complex processes
• Incompatible legacy systems

As a result, your own employees are transformed from trained specialists into technology administrators.

Typical automation failures using the example of supplementary dental insurance

Let's look at an example: The insurance company we have chosen is a private insurer that has to deal with millions of claims from its policyholders every year. This includes, among other things, reimbursing complex dental treatments under supplementary insurance. The insurance company uses a variety of technologies:

a digitisation pipeline that converts paper documents into digitally readable documents, generative AI for processing all findings, and a CRM system, which is, however, already quite outdated. In addition, service providers must be verified via an online portal. The foundations for full automation appear solid, but we encounter significant problems in implementation:

• The current rigid process requires documents to be submitted at the beginning of the process. Documents that need to be processed in the middle of the process, for example when insured persons submit additional information, cannot be passed on to the digitisation pipeline.
• Although the AI delivers very good results, it is not integrated into the process flow at all.
• The CRM system is outdated and does not offer any interfaces.
• The online platform is also optimised for human use and offers us only a few technological interfaces.

The technology should fit the process, not the other way around

So should the idea of full automation be discarded? One option would be to replace all technologies with a new platform. This would be a costly and time-consuming undertaking, the result of which would have to address the question of suitable, future-proof interfaces in parallel with the existing online platform.

Our goal: business processes must be illustrated in such a way that different technologies can be used as needed and according to the task at hand. Missing interfaces must be bridged. The use of technology must be process-oriented, not the other way around.

Intelligent use of technology and full automation with workflow automation

The solution is an intelligent combination of workflow management systems based on BPMN 2.0 (Business Process Modelling Notation) and Robotic Process Automation (RPA). The resulting workflow automation is a groundbreaking step towards a fully automated, digital world.

Let's take a closer look at these technologies and how they work together as workflow automation:

The technology integrator: workflow management system and BPMN 2.0

Common workflow management systems take a holistic approach to representing business processes. BPMN 2.0 allows complex business processes to be mapped using activities, conditions and sub-processes in a low-code approach. The modelling is by no means purely descriptive. Specific requests and process cases are fed into the process model, enriched with information and processed. A customer request is like a token that processes the described stations according to the defined logic and thus runs through the model via the desired process path. At the individual stages, various technologies can be accessed via an API, allowing information to be sent or received. This is where the biggest advantage and disadvantage of using a workflow platform in isolation becomes apparent. If technologies or systems without interfaces are accessed during the process, automation is not possible. A human must be involved, which results in a process break.

The data broker: RPA

This is where RPA comes into play as a data broker and technology bridge. Robotic process automation uses software robots that communicate and interact with software via the graphical user interface (GUI), thereby automating processes and process sequences where classic integrations fail. As a rule, RPA developments – which are also developed using a low-code approach – are used in small process sections where high-frequency, strict sequences can quickly create economic added value. However, they are unsuitable for automating extensive, usually time-consuming business processes.

The application of workflow automation

By combining the two technologies and using workflow automation, we can solve the problems of private insurance. The low-code approach of the workflow platform maps the business process with all its logic, branches and stages. Different document inputs – for example, the subsequent submission of missing documents – are no longer a problem, as we can control them in an orderly manner within the process flow by modelling the process in BPMN 2.0. Both the digitisation pipeline and AI offer us interfaces that can be accessed via API. When the appropriate technologies are used, the information is transferred to the software at the respective process step and the result is awaited. For each process step in which the CRM system is to be addressed, RPA takes over communication and interaction, thus bridging the missing interfaces with the help of the GUI. In this way, complete automation is achieved through workflow automation.

The adesso advantage

At adesso, we are intensively engaged with precisely this form of automation. Years of experience in working with various automation technologies have shown us how business processes can be automated sustainably, efficiently and profitably. The solution does not lie in a single technology, but in recognising the capabilities of each individual technology, exploiting its strengths and weaknesses, and bringing them together in a scalable platform in a task-oriented manner: workflow automation.