ERP Data Migration to SAP and COUPA
A data migration program transforming data from multiple ERPs into SAP and COUPA target systems.
Business issue
Why I was brought into the project
GEODIS needed to migrate data from multiple ERP systems into SAP and COUPA while preserving quality, structure and alignment with target-system requirements.
Context
The environment around the project
A logistics group needed to migrate data from multiple ERPs to SAP and COUPA.
Functional environment
The functional context covered ERP migration, procurement processes and business data alignment with SAP and COUPA.
Technical environment
The technical environment combined Dataiku, PySpark, Impala, HDFS, Hive, Azure, Nifi, Kafka, Airflow and Starburst.
Challenges
The project required adapting heterogeneous source data, building high-performance transformations and coordinating with both technical and functional experts.
Solution
My contribution and its impact
My contribution to the project
I designed and deployed scalable transformation processes in Dataiku and PySpark, adapting data to target-system requirements and improving migration execution.
- Scalable migration transformation pipelines
- Data adaptation logic for SAP and COUPA
- Performance-optimized processing flows
- Reusable Dataiku migration assets
Impact
The project helped make a complex migration more controlled by turning heterogeneous data preparation into repeatable, optimized pipelines.
- More reliable preparation of migration data
- Better alignment with target-system constraints
- Improved scalability of transformation processes
Impact metrics
Approach
How I structured the work
- Analyze source ERP structures and target SAP and COUPA requirements.
- Build optimized transformation pipelines using Dataiku, PySpark and Impala.
- Iterate with technical and functional experts to validate migration readiness.
Takeaways
What I learned from this project
- Migration success depends as much on functional alignment as on technical transformation.
- Performance matters early when data volumes and repeated migration runs increase.
- Dataiku can structure collaboration between data engineers and functional experts.