Associate Consultant at SplashBI

I built query-based enterprise reports over Oracle EBS Procure-to-Pay and Order-to-Cash data, writing SQL and PL/SQL against schemas that had accumulated years of special cases.

The largest piece of work was a migration. Between 25 and 50 legacy Discoverer reports had to move through EIS Explorer onto the SplashBI platform. Porting them was straightforward. Proving the numbers still matched was not, and that is where most of the time went: every report was validated across all three systems before it was signed off. A reporting migration that quietly changes a total does more damage than one that fails outright, because nobody goes looking for it.

I also wrote RESTful APIs in Node.js for blog management workflows, covering CRUD operations and integration with the existing backend services.

What I took from it#

Enterprise data taught me that correctness is partly a social question. The schema tells you what values are permitted. Only the people who use the report every month can tell you what those values are supposed to mean, and they are usually the ones who notice when you get it wrong.