Legacy ERP conversion, rebuilt on Azure

A single-machine ERP conversion workflow at its limits.

Problem

A single-machine ERP conversion workflow at its limits.

Approach

An aging Enertia ERP conversion system - the workflow that imports and updates client company data into the ERP - was rebuilt on a hybrid-cloud approach I co-designed. We uplifted the conversion workflow onto newly architected Azure data infrastructure: Azure Data Factory for orchestration, Azure Databricks with PySpark for transformation at scale. The result: a legacy, single-machine conversion process became a governed cloud pipeline able to onboard data from multiple client companies into the ERP reliably and repeatably.

Architecture

Legacy conversion single machine ADF + Databricks PySpark at scale ERP multi-company onboarding

Modernizing without breaking the business

The riskiest system to modernize is the one quietly feeding your ERP. The hybrid approach kept the legacy path alive while the Azure pipeline proved itself in parallel - conversion by conversion - until cutover was a non-event. That's the pattern for any legacy data workflow: parallel run, parity checks, then retire.

Stack

Have a similar problem? Let's catch up →