About me2026

Marcus Belz

Senior Database Developer

25 years of database development - domain depth in the reference projects, hands-on technical depth in the DI² side project.

View projects DI² side project

25y

DB development 1997 - today

12y

Years ETL focus since 2014

Industries CRM to geodata

3TB

Largest migration ~3 TB · 150 IF

How I workthree sentences

ETL rarely fails on technology. It fails on missing understanding.

Understand the data before you process it.

Read the data model, interview the source systems, document the domain quirks. "What does NULL mean in this column?" decides whether a migration project succeeds.

Domain-first · not tool-first

Transparency over processes and data quality.

Three logging levels - process, step, stack trace - queryable with a SELECT. Faulty records land in an error table, not silently in the warehouse.

Queryable · traceable · auditable

Robust processes and maintainable code.

Generic pipeline generation from metadata. One procedure per SQL file, every code change visible as a Git diff. Code someone can still read 18 months later.

Maintainable · readable · repeatable

Industry knowledge5 worlds

Data understood
across 05
industries.

Every industry has its own data models, its own pain points, its own stakeholders.

What stays the same: the same discipline for robust, maintainable ETL pipelines.

01
DI² - side project
di2.marcus-belz.de
ETL generator from metadata - maintainable, readable, auditable. Plain SQL as output, runnable in any SQL engine. Own infrastructure on Hetzner, AI-driven development workflow.
2026 - ongoing
Next.js 16PostgreSQL 17KeycloakDocker
02
Geodata & PostGIS
GIS sector
Architect for database design and development of geodata in Postgres/PostGIS and SQL Server. Spatial analysis and data processing server-side in the database.
since 2023
PL/pgSQLPostGISPython
03
Industry 4.0 & steel alloys
adesso SE · steel/specialty metals
Performance-critical access to the control systems of melting furnaces and rolling mills. CDC against Oracle, Data Vault historization, ~500 tables from 6 plants.
2021 - 2022
CDCSQL ServerT-SQLOracleData Vault
04
Automotive & logistics
Daimler TSS · automotive logistics
Controlling of transport capacities and damage management. Reconciling planned against actual routes, damage management for carriers, vehicles and routes - complex topological logic.
2016 - 2018
T-SQLSSISQlik Sense
05
Data migration & Microsoft Dynamics 365
Infoman AG · sanitary/building supplier
Group-wide consolidation: 30 Lotus Notes apps → Microsoft Dynamics 365. 150 interfaces, 100,000+ XML files, delta migration via hash. ~3 TB data volume.
2014 - 2016
T-SQLSSISDynamics CRMAzureCozyROC

Tech stack41 tools

marks core competencies

What
actually
runs.

No buzzword bingo. This is the stack the five projects above were actually built with - from source-system analysis to the production pipeline.

databases

PostgreSQL
SQL Server
PL/pgSQL
T-SQL
PostGIS
Dynamic SQL
Oracle

etl

Kimball/Inmon
Data Vault
SCD2
Hash-Delta
CDC
SSIS
Talend
CozyROC
Scribe

formats

XML
JSON
CSV / Flat
Webservice / REST
Bulk-Request
Lotus Notes
SharePoint

frontend

Power BI
Qlik / QlikView
Next.js 16
React 19
TypeScript
Tailwind

coding

C#.NET
VB.NET
Python

infra

Docker
Compose
GitHub
Liquibase
Keycloak
Nginx

Substance over marketing.

Projects · selection

Reference projects in detail

CRM mega-migration · logistics EAI · Industry 4.0 · geodata - task, approach, outcome.

Project overview

Side project · di2.marcus-belz.de

DI² - ETL generator

Vision, infrastructure, development workflow - how I build my own product in a structured way, from spec through UX to deploy.

DI² in detail

Get in touch.

Questions about data migration, ETL standards, warehouse architecture or DI² - happy to talk by e-mail or through the usual networks.

Write an e-mail LinkedIn XING