★ Where it started

ERP
Analytics

Oracle Cloud HCM · Bangalore · Aug 2017 - Jul 2019

Q4 forecast NA+12%

Q4 forecast EMEA+4%

Q4 forecast APAC+9%

SaaS attach rate+2 pp

Project Overview

My first job out of grad school. Two years across the full app lifecycle on Oracle Cloud HCM - integration, automation, ML for forecasting, and a credit-card fraud POC on the side. The work that taught me production systems aren't just code, they're paperwork.

Oracle Cloud HCMJava · PythonSQL · NoSQLELK StackAWS KinesisKafka50K+ LOC13 business units

The work

PIP automation. Automated the Performance Improvement Plan process across 13 business units. Saved 600+ monthly work hours.
ERP analytics dashboard.Real-time overview for the client CEO with quarterly ML forecasts. Tied to a 20% quarterly business activity lift.
Fraud detection POC. AWS Kinesis + Kafka pipeline for real-time credit-card classification based on user buying patterns.

600 hrs

Monthly work-hours saved by automating the Performance Improvement Plan process across 13 business units.

2× faster

Custom report generation

2 months

Project delivered early

The system, in one diagram.

Four layers: Oracle Cloud ERP application sources, Kafka & Logstash ingestion bus, real-time Spark streaming & fraud scoring, and Elasticsearch/Kibana SOX analytics.

Oracle Cloud ERP Analytics and Real-Time Fraud Detection Architecture: ERP Sources, Kafka Ingestion Bus, Spark Streaming Aggregations, and Elasticsearch Kibana Dashboards

The outcomes.

+20%

Quarterly activity

Tied directly to insights surfaced by the new analytics dashboard.

95%

Test pass-rate

Test scenarios and cases authored for HCM integration releases.

2 POCs

Pipelines shipped

ELK for Spark logs + Kinesis/Kafka for fraud-detection - both adopted by the team.

ERPAnalytics

Business units