# Who I am

> Ten years modeling, integrating and making data reliable — from legacy BI to the open source Modern Data Stack. On-premise, open source, European sovereignty.

- URL : https://blog.xreveillon.eu/en-US/about
- Date : 2026-07-17T00:00:00.000Z
- Langue : en-US

---

Ten years modeling, integrating and making data reliable for banking, insurance and the public sector — first on the BI and legacy ETL side (Informatica, Talend, ODI, SSIS), now pivoting to the **Modern Data Stack** open source: Dagster, dbt, Airbyte, Parquet/Iceberg, DuckDB, MinIO/S3.

Member of the **Edda AI team**, in charge of determining what we *can* do with AI, what we are *allowed* to do, and how to bring it concretely into working groups — not drop it in as a black box.

## The data model first

Conviction: **the data model first**. Medallion architecture, idempotence and replay at the core of design, observability and specifications before code. On-premise, open source, European sovereignty.

## Homelab since 2015

Family self-hosting since 2015: personal data platform (Kubernetes, MinIO/S3, Dagster, Iceberg, OpenMetadata), mail server, Nextcloud, Immich, Bitwarden, private PKI, monitoring, home automation. It's the practical root of my entire on-prem and open source philosophy — what I advocate at work, I run at home.

## Teaching & sharing

I run internal knowledge-sharing sessions and present technical topics at Edda. I mentor juniors (Edda colleagues, Telecom Nancy alumni network). Teaching and sharing are active soft skills for me: explaining technical concepts to non-technical people matters as much as mastering them.

## Watchlist

Modern data stack, data quality and governance, AI and sovereign agents, software craft, European digital sovereignty.