// Project

SocGen GTS Alerting

[ Context ]

Societe Generale's GTS Alerting team needed to stabilize and scale the monitoring pipeline for their private cloud, which was suffering from production crashes and inconsistent data source handling.

[ Objective ]

Fix critical production bugs, redesign the core Java microservice around Apache Flink, and establish reliable CI/CD practices across environments.

[ Approach ]

Re-engineered the Flink-based microservice at the core of the monitoring stack. Designed reusable abstractions for Flink data sources. Collected custom metrics sent to InfluxDB and built live monitoring dashboards in Grafana. Advocated for and implemented GitOps-based CI/CD with Jenkins.

[ Outcome ]

Stabilized a system processing 14.5B daily events. Eliminated recurring production crashes and established consistent deployment practices across environments.