Internship series 1: Godfred and Thimo

Introduction 

We are Godfred Mensah and Thimo Van Reeth, two students in Applied Computer Science at HOGENT, specializing in Data & AI. While looking for an interesting internship, we found Alluvion’s assignment through the HOGENT internship app. The combination of SAP, artificial intelligence, and automated documentation immediately caught our attention. That is why we decided to apply for this internship assignment at Alluvion. 

During our internship at Alluvion, we had the opportunity to work together on an assignment within the SAP environment: using artificial intelligence to automatically document ABAP code.

Onboarding and First Steps at Alluvion 

Our internship started with a two-week bootcamp. During the first week, we were introduced to the internship assignment and started with several SAP learning modules. These included topics such as “What is SAP?”, discovering SAP, and an introduction to MDG. This gave us the basic SAP knowledge we needed before starting the actual project. 

In the second week, the focus shifted more towards MDG. We attended an MDG inspirational session and an Alluvion pitch given by Wouter Van Peteghem. After that, we followed several MDG Admin Training sessions. These sessions were not only theoretical but also included practical and technical exercises in the SAP system. This allowed us to try out different MDG concepts ourselves and understand how they are applied in a real SAP environment.  

Besides the learning sessions, the onboarding period also included an informal dinner with the other interns and mentors. This was a nice opportunity to get to know everyone in a more relaxed setting and helped us feel more connected to the team from the beginning.

Our Internship Assignment 

Our internship assignment focused on the automatic generation of documentation for SAP code. In many SAP environments, documentation is important, but creating and maintaining it often takes a lot of time. Code changes regularly, while documentation does not always evolve at the same pace. Therefore, we investigated how AI can help analyse ABAP code from an SAP system and turn it into clear technical and functional documentation. 

The goal of our assignment was to develop a proof of concept. This proof of concept had to show how SAP code can be extracted, analyzed and automatically documented. The focus was not only on the technical side, but also on the usability of the output. The generated documentation had to be understandable for developers, consultants and other colleagues who work with SAP.

Our Approach 

To bring this assignment to a successful end, we worked step by step. First, we familiarized ourselves with the SAP environment and the way ABAP code is structured. Afterwards, we investigated different ways to extract code from SAP, for example through APIs, existing SAP tools or other technical approaches. To conclude, we explored how AI models could be used to understand the structure, logic and purpose of the code.

Our Experience at Alluvion 

Our experience at Alluvion was very positive. From the start, we were given trust and room to work independently. At the same time, support was always available when needed. There was room to ask questions, discuss ideas and explore new technologies.

Guidance from Our Mentor  

The guidance from our mentor, Dylan Van Paemel, was also very valuable. He helped us think through the technical approach, gave feedback on our ideas and supported us whenever we got stuck. What we appreciated most was that we were not simply given ready-made answers. Instead, we were encouraged to think for ourselves and to support our decisions with clear reasoning. This helped us become more independent and gave us a better understanding of how technical decisions are made in a business context.