How to start with AI in insurance, practical guide (part 3.)

At the request of VVP, featured in  the latest issue  VVP, the leading platform for financial advisors in the Netherlands, our CTO and AI strategist, Dennie van den Biggelaar, explains how to apply specific AI and machine learning to ‘advice in practice.’ In various editions, the following topics will be highlighted:

  • Starting with specific AI and ML
  • Operationalizing in business processes
  • Integrating into existing IT landscapes
  • Measuring = learning: KPIs for ML
  • Ethics, regulations, and society
  • AI and ML: a glimpse into the near future

In this third edition, we answer the question: how do you integrate a trained algorithm with your existing IT landscape and tooling?

Integrating AI Software

The insurance sector is on the verge of a technological revolution. With the integration of AI decision engines, insurers can significantly improve customer service and achieve better business results. AI algorithms can predict churn, calculate customer lifetime value (CLV), and make recommendations for cross- and upsell, enabling advisors to make better-informed decisions. But how do you integrate these algorithms into your existing IT landscape? How do you ensure that your employees have these predictions and suggestions at the right time to work more easily and effectively? This article discusses several concrete technical tips for successfully integrating AI decision engines into insurance systems.

Define a Successful Integration

I firmly believe that IT issues must always serve a business goal. A successful integration always begins with the question: ‘When is this integration successful?’ Creating a user story can help, for example:

“As [digital marketer at underwriting agency X], I want to [know weekly which customers need an additional product Y], so that I can [set up a targeted automatic marketing campaign for this group] with the aim of [generating (new) leads weekly for my field advisors].”

This is a good starting point to present to technical experts what is expected of them. Usually, further questions follow:

  • What specific information does the user want to see?
  • How often should it be refreshed?
  • How will we measure the success of these automatic campaigns?

By asking and answering these questions, the team naturally identifies the framework of a successful integration. This is not a one-man job: it is important that both business/users and technical experts are represented in this exercise!

Analyze Existing IT Landscape

A successful integration begins with a thorough analysis of the existing IT infrastructure. Many integration attempts fail due to a lack of understanding of current systems, leading to compatibility issues. With which existing IT systems, databases, and interfaces must the AI algorithm ‘collaborate’? What amounts of data need to be transferred? When and how quickly?

(Successful integration means creating synergy between backend and frontend systems.)

In practice, this means working and aligning with various IT partners of backend and frontend systems. Start this inventory early and include all (external) stakeholders in your plans. If you don’t have the time or resources for this yourself, appoint one of your IT partners to do this project management for you. After all, it is their expertise!

Clear, Scalable, and Agile

Unfortunately, I have often seen organizations with innovative plans whose IT landscape was too rigid. Therefore, design a modular and scalable architecture to allow for future expansions and changes, ensuring your organization remains agile. Nowadays, it is best practice to use microservices architectures, where each functionality runs as a separate service. This makes it easier to add, replace, or update new elements without overhauling the entire infrastructure.

Consistency and Quality

Data quality is crucial for the success of AI. Many AI systems perform poorly due to inconsistent, incomplete, or outdated data. Therefore, implement a data cleaning and preprocessing pipeline to ensure that all data sent to the AI decision engine is clean and up-to-date. Automated tools for data integration and validation can help ensure the reliability of AI outcomes. Use ETL (Extract, Transform, Load) processes to extract data from various sources, transform it into a uniform format, and load it into a central data repository. This ensures a streamlined data flow essential for successfully training and using AI models.

Testing, Validating, and Monitoring

Thorough testing and validation are essential to ensure that AI models function correctly within existing systems. Insufficient testing can lead to errors and unexpected problems after going live. Therefore, conduct extensive tests in a simulated environment that mimics the production environment. Validate the output of AI models with historical data and scenario analyses. Involve end-users in the testing phase to ensure that the models meet business requirements and user needs.

Use APIs

APIs (Application Programming Interfaces) are essential to connect AI decision engines with existing systems. Without standardized interfaces, communication between systems can be inefficient and problematic. By developing and implementing APIs that can receive and send data, integration becomes flexible and scalable. This ensures that the AI decision engine can robustly communicate with both back-office and front-office systems.

Security and Privacy

Data security is crucial, especially given the sensitive nature of insurance data. Insufficient security can lead to data breaches, resulting in loss of customer trust and privacy violations. Therefore, use only the data that is truly necessary and anonymize as much as possible. If you really need certain sensitive data, apply encryption. Ensure that all data transfers between systems and the AI decision engine run over an encrypted connection. Use access controls and audit logging to guarantee data security and compliance with regulations.


A thorough integration is a prerequisite for AI to land successfully in your organization. You must consider not only a scalable business case and the user but also think about agility, security, privacy, and quality. Therefore, you need a team with different competencies and must coordinate and collaborate with IT partners.

First, ensure that you or someone in your organization has a clear understanding of the business framework of a successful integration. Make this explicit so that you can convey it. Then appoint someone responsible for the realization and the associated project management. If you do not want or cannot free up resources for this, you can easily engage one of your trusted IT partners for this. You can then focus on your own core business!

“It’s easy to create a self-learning algorithm. What’s challenging is to create a self-learning organization.” – Satya Nadella, CEO of Microsoft

In short: devising, building, and validating a robust algorithm is only phase one of successfully implementing AI in practice. In the next edition, we will delve into how to integrate it with existing IT systems and workflows.

Post Tags :

Share :

Scroll to Top