The “headless” learning platform – an integral part of Union Investment's advisor portal

Union Investment is a German fund management company that is part of the cooperative financial group (including Volksbanken and Raiffeisenbanken) and currently manages €534.6 billion in client assets. It offers investment funds and other investment products for private and institutional investors and serves a total of more than 6 million clients.

The Challenge

Union Investment’s core target group consists of the financial advisors at the various Volks- and Raiffeisenbanks. They require extensive product knowledge, access to specialized information, and, in particular, an understanding of new and complex investment funds and investment products. In addition to their initial training, financial advisors are required to regularly participate in training and continuing education courses. These programs cover various areas of financial consulting and ensure that advisors are always informed about current developments and legal changes.

To this end, Union Investment planned the development of a multifunctional fincancial advisor portal called “InvestmentWelt.” It brings together all functional areas—from pure information and specialized applications such as calculators to the topic of learning—within a consistent and unified user interface. Against this backdrop, the fund company was looking for a learning platform that could be flexibly adapted to its individual needs and integrated seamlessly with the other components of the portal.

Furthermore, the solution naturally had to meet the high IT security requirements of a financial institution and was also intended to be operated together with the other components in a unified system architecture in the Azure Cloud.

The Solution

Totara emerged as the ideal solution for this challenge: Totara’s customizable architecture provides a flexible foundation for delivering a tailored learning platform to the client. Union Investment’s evaluation revealed that Totara best met the requirement for seamless front-end integration.

Another key deciding factor was the ability to run the platform within the client’s own Azure cloud infrastructure. It turned out, however, that there was very little experience with hosting Totara on Microsoft Azure. For this reason, we worked closely with the DevOps teams to develop and implement a suitable operational architecture for Azure. After several weeks of evaluation and technical testing, Totara was successfully deployed in the Azure environment.

Architectural Concept

The platform’s architecture is based on the principle of a Self-Contained System (SCS). Under this approach, the learning platform is built and operated as a standalone, independent system and contains all the components, resources, and information within Totara that are necessary for learning. This learning component is then integrated into the portal via API in a “headless” manner. The goal of this architectural approach is, on the one hand, to keep all relevant learning logic within Totara: This ensures processes that are as simple as possible, improved maintainability, and flexible scalability. On the other hand, it also enables completely seamless integration in the frontend. The otherwise often noticeable and perceptible breaks in the user journey—e.g., between a CMS and an LMS—could thus be completely avoided.

API Strategy and Extensions

Totara’s GraphQL API was used for the integration and further development of the platform. Since some required functions were not included in the standard scope, a custom extension plugin was developed. This supplements the existing GraphQL API with project-specific endpoints and functions.

One advantage of this approach is that individual extensions can be managed separately and removed at any time. This ensures that the learning platform remains fully updatable.

Special Plugins & Integration Solutions

A custom plugin was developed to extend activity access restrictions. This plugin allows certificates to be issued only after a defined period of time has elapsed following the completion of an activity.

Another plugin was used to create an additional learning activity for integrating videos from the MovingImage system (VideoManager Pro). Editors can create video activities within Totara and store the respective video ID.

Authentication and User Management

Login is handled via Microsoft Entra ID. To enhance security, authentication based on JSON Web Tokens (JWT) was additionally implemented. A dedicated authentication endpoint was developed on the Totara side for this purpose. JWT enables secure and standardized transmission of user information in accordance with RFC 7519.

Target Group Assignment

Target groups are assigned using two mechanisms:

1. During login, the portal reads target group information from a JWT and stores it in the Totara user profile. This information is then available for the target group-specific delivery of courses.
2. Additionally, a background process ensures that target group assignments are also synchronized for users who have not logged in for an extended period. This keeps the data consistent and up-to-date.

Access to Learning Resources

The frontend requires access to resources such as images, documents, or SCORM packages located within the Totara environment, which is accessible only internally. API calls are handled via Azure API Management (APIM). However, a reverse proxy approach was deliberately chosen for accessing resources. Resources are therefore not served directly via Azure APIM, but delivered via a dedicated reverse proxy. This has enabled the creation of a high-performance and secure access solution for static content.

Elasticsearch Integration

The “InvestmentWelt” portal provides a global search function based on Elasticsearch. To achieve this, a custom integration solution was implemented to connect Totara courses and content to the portal’s global Elasticsearch database.

An event observer monitors changes to courses and responds to the creation, update, or deletion of learning content. The corresponding data is then automatically indexed in Elasticsearch or removed from it.

Additionally, the search logic takes into account course publication and expiration dates:

• Courses only appear in the search results starting from the defined publication date.
• Once the expiration date has passed, courses are automatically removed from the search results.
• Archived courses must have an expiration date so that they do not continue to appear in search results

The Result

Today, Totara serves as the central learning platform for the “InvestmentWelt” portal. The portal itself enjoys great popularity thanks to its optimized user interface and the integration of numerous value-added services. With almost 40,000 active users, the learning platform is a central component of the portal.

Currently, 25 fixed learning programs, including regular reporting, are being run on the platform. As part of an evaluation of ten learning programs, enrollments and completions from March 2023 using the old LMS were compared with the figures from March 2024 following the launch of the new portal. This revealed a 68% increase in enrollments and a 65% increase in completions.

The learning platform was implemented over a period of approximately 1.5 years leading up to go-live. In parallel, four to five additional teams worked on adjacent systems and business areas within “InvestmentWelt.” These teams were coordinated through agile project management based on SCRUM. The platform continues to be developed even after going live.

A key component of the project was the implementation of security and data protection requirements. In addition to the requirements of the GDPR, the regulatory requirements of the financial sector also had to be taken into account. External consultants can access the portal, while the actual Totara platform is operated within a secure cloud environment at Union Investment that is accessible only internally. As a result, all audits were successfully completed, and a clear separation between external access and the internal system landscape was ensured.

The platform is not used exclusively for traditional learning content but is also utilized for surveys, studies, and testing formats. To date, approximately 8,500 participants have been recorded. The collected data supports the identification of product improvements and organizational measures.

In addition, recurring webinars and events organized on short notice are managed via the platform. Event registrations, in particular, sometimes generate high traffic volumes with several thousand concurrent requests, which demonstrate the performance and scalability of the infrastructure and the architectural approach to integration.