Community Meetup @ InfrabelJoin our next meetup on June 25th! We will have 3 talks and will share some food and drinks while networking
**Location:** Infrabel, Rue des Deux Gares 82, 1070 Anderlecht, Belgium
**Agenda:**
17:45 : Doors open
18:00 - 18:40 : Drinks and pizza
18:40 - 18:45 : Welcome
18:45 - 19:00 : Elastic @ Infrabel
19:00 - 19:15 : Real life never-seen-before railway project
19:15 - 19:55 : From unknown and raw to real value
19:55 - 20.20 : What’s new in Elastic v9
20:15 : Wrap up
**Talks:**
**Elastic @ Infrabel**
In this presentation, we will explain how Infrabel uses Elastic today and how it supports our current needs.
We will briefly outline some future optimizations and the integration of OpenTelemetry to further strengthen and extend our Elastic platform.
To showcase how we use Elastic for addressing real business use case, we will demo a never-seen-before application build in collaboration with Elastic Belgium.
*Safia Souabi – Big Data Engineer at Infrabel*
*Olivier-Désiré Muhimpundu – Big Data Engineer at Infrabel*
**Real life never-seen-before railway project**
*How Elastic helps ensure the punctuality of rail services by detecting and correlating events in real time to anticipate disruptions, ensure operational continuity and optimise the flow of rail traffic."*
*Vanessa Coeurnelle – Cyber Resilience Data Analyst at Infrabel*
**From unknown and raw to real value**
In this talk, Chris shares how he transformed a legacy log ingestion pipeline into a structured, scalable, and insight-ready ELK-based system. Starting from loosely indexed XML/SOAP logs with minimal configuration, the project required rebuilding the data model to handle multiple message types, complex nested structures, and inconsistent field representations.
Without the luxury of a full upfront redesign, Chris adopted an iterative, data-driven approach: combining exploratory analysis with dynamic mapping strategies to progressively shape the data into a usable format. He will discuss how this led to the design of flexible ingestion pipelines, including XML parsing, normalization, and generic handling of heterogeneous payloads.
The talk also covers practical techniques for managing sensitive data, reducing storage overhead, and improving query performance—such as selective data retention, hashing and fingerprinting strategies, and the use of data streams. Finally, Chris will touch on how structuring the data effectively enables higher-level capabilities like anomaly detection and automated alerting.
This session is aimed at practitioners dealing with messy, real-world data and looking for pragmatic ways to make Elasticsearch pipelines more robust, maintainable, and valuable.
**Bios**
**Christoph Evers** is a Senior Data Scientist and Data Engineer based in Brussels, with over 15 years of experience in data systems, search technologies, and analytics. He has been working with Elasticsearch since its early days, contributing to the project between 2010 and 2012 and applying it extensively as a search engine, document store, and log ingestion platform.
At OpenLex, he designs and builds advanced data solutions for legal research, including full-text search engines, NLP pipelines for entity detection and anonymization, and large-scale analytics systems. His work also includes operating and optimizing ELK-based observability stacks in Linux and Docker environments.
Previously, Chris held roles spanning software engineering, data science, and R&D leadership across sectors such as social media analytics, mobility and market research. His experience combines hands-on engineering with business-oriented data applications, delivering scalable systems and actionable insights for both startups and multinational clients. He mainly focuses on search relevance, data observability, and the practical use of ELK in production environments.
**Olivier-Désiré Muhimpundu** is a Data Engineer at Infrabel, where he has been working for more than three years on large-scale data platforms and data projects. He holds an engineering degree from University of Mons Faculty of Engineering and a Master's degree in Artificial Intelligence.
In his current role, Olivier focuses on designing, operating, and evolving enterprise data platforms based on technologies such as Apache Kafka, Elastic Stack, and Cloudera. His work spans data ingestion, real-time processing, monitoring, observability, and platform engineering in a large-scale railway environment.
**Safia Souabi** is a Data Engineer with 6 years of experience at Infrabel in Brussels. With a BSc and MSc in Physics from London, she combines a strong analytical background with expertise in building and operating large-scale data platforms for mission-critical operational and asset data. Experienced with Apache Kafka, Elastic Stack, and the Cloudera ecosystem (Hadoop, Spark, Hive/Impala), she delivers scalable batch/real-time data pipelines and optimized ETL/ELT workflows. She works closely with stakeholders to translate requirements into reliable data products, with a strong focus on data quality, governance, security, performance, and platform resilience.