ETL & Data Warehouse Consultant
Most reporting problems are data-plumbing problems. I build the pipelines and warehouse that turn scattered ERP, warehouse, point-of-sale and spreadsheet data into one trusted source, so every dashboard agrees with every other one.
What a warehouse engagement covers
Source analysis (SAP Business One on HANA or SQL Server, INFOR WMS, POS systems, MySQL applications, Excel), a dimensional model designed around the questions the business asks, ETL pipelines in SSIS, Talend or SQL that load incrementally and log what they did, data-quality checks, and a Power BI layer on top. I currently run exactly this as Software Team Lead at Oak Tree Ventures Group, centralising SAP Business One, WMS, Distech and Order@ data into a unified warehouse.
Data audits
Before building anything, I can audit what you have: duplicate customers, inconsistent KPIs, unclear ownership, reports that disagree. You get a written findings report with a prioritised fix list. Many clients start here.
What's included.
Dimensional modelling
Star schemas with conformed dimensions so sales, inventory and finance reconcile.
ETL pipelines
SSIS, Talend and SQL-based pipelines with incremental loads, logging and restartability.
SQL Server & SAP HANA
Staging and warehouse design on MS SQL Server, sourcing from HANA, MySQL and ODBC systems.
Data quality
Validation rules, reconciliation against source totals, and alerts when loads fail.
Integrations
WMS, POS, e-commerce and IoT feeds landed in the warehouse through SSIS and APIs.
KPI & process design
New KPIs and processes for commercial and 3PL operations, defined once and reused everywhere.
Questions,
answered.
Do we need a data warehouse, or is Power BI enough?
If you have one clean source and a small team, Power BI alone is often enough. Once you have several systems, history that needs preserving, or refreshes that take hours, a small warehouse pays for itself quickly.
Which platforms do you use?
Mostly Microsoft SQL Server with SSIS, plus Talend and plain SQL. Sources include SAP HANA, MySQL, INFOR WMS and ODBC-based systems.
How long does a first warehouse take?
A first version covering one or two sources typically ships in six to ten weeks, with further sources added incrementally.