porta holding Integration of structured and semi-structured master data

porta holding
Customer
porta Logo

porta
Industry: Furniture retail
Headquarters: Porta Westfalica

Your contact
Rouven Homann, Member of the Board, cimt Group
Rouven HomannMember of the Board, cimt Group
Consulting focus
Quantifiable benefits

The customer

The porta Group currently comprises 27 large porta furniture stores, two porta kitchen showrooms in Berlin-Mahlsdorf and Leipzig-Paunsdorf, around 100 stores of the omnichannel furniture discounter SB-Möbel BOSS, 22 furniture stores of the ASKO Group in the Czech Republic and Slovakia, and the Saxony-Anhalt furniture retailer Möbel Letz. The group, with around 7,000 employees, generated annual revenue of EUR 1.34 billion in 2022.

Initial project situation

The project started at the beginning of 2023. In parallel with the introduction of a modern PIM solution in AWS, the aim was to build an integration platform with Talend to transfer product data from the master data systems of several companies to the multibrand storefront with the shortest possible latency. 

The multibrand storefront in GCP was also being developed in parallel by several development teams at porta. It was therefore clear from the outset that close exchange between both parties was essential. 

Greenfield

The project started as a greenfield project. Together with a small team of developers and leads from cimt and developers from porta, we got the project off the ground. The initial focus was on setting up the environment for the Talend infrastructure and training porta’s developers. The training covered Talend basics and advanced content as well as API training. We also agreed on an agile development process with which the developers on both sides could work as quickly and efficiently as possible. The choice fell on a healthy mix of Scrum and Kanban, which suited the team size and goal orientation perfectly. 

After the first few iterations, a CI/CD pipeline was also implemented with the help of the cimt DevOps team. The DI and API applications were ultimately deployed in a GCP environment. 

Target architecture

One of the main goals was to reduce the latency between changes to a minimum so that changes in the PIM system become visible in the web shop much faster. To achieve this, the PIM system notifies Talend on an event basis as soon as changes have occurred. However, the event itself only contains the reference, so Talend then has to retrieve and process the detailed data. 

The web shop’s REST API is then used to transfer the data to the target and supply the multibrand storefront. 

porta – Talend-Integration: PIM, Talend, GCP, Multibrand Storefront

Data mapping

Since we are talking about a multibrand storefront in different languages, we agreed on a generic approach to mapping. The meta-information of the mapping was stored in the database, and the actual mapping was particularly easy to implement with the Talend Data Mapper.

In other words: source and target properties were persisted in a database together with transformation information, and the Talend Data Mapper used this information to transform the data from the PIM. 

Features and frameworks for the implementation

The cimt JobInstance Framework (JIF) together with its Talend custom components is dthe framework of choice for dthe use of Talend DI applications. The JIF enables developers to establish delta processing more quickly , supports monitoring and provides lineage information – to name just a few of its features. 

Another noteworthy feature is the GCP Secret Manager, which stored all credentials and connection parameters (host, port, etc.) for all services . If services moved or credentials were rotated, only a change to the secret was necessary, as the Talend jobs retrieve the information independently. 

Why cimt?

Thanks to cimt’s many years of experience in software development and data engineering projects, we knew we had a good partner at our side – not only to introduce Talend ourselves, but also to achieve our ambitious goal. 

Porta IT has taken a step towards modern development together with cimt, and we look forward to further joint projects. 

“The seamless integration of the cimt consultants into our team was remarkable. Such a close and effective partnership developed that it felt as if our colleagues from cimt had been part of our company from the very beginning. Their professionalism and reliability convinced us, and we look forward to achieving further successes together.” 

IT team – Integration & Automation porta IT-Service GmbH & Co. KG

References

More business cases

GEMA – New data strategy with Databricks

GEMA – New data strategy with Databricks

GEMA used SQL, Excel and manual reconciliation for its reporting. cimt supported the build-up of a lakehouse platform with Databricks, through which business units now receive their data automatically – in under two hours instead of two to three days.

Would you like to find out more?

Scroll to Top