
OOMKilled Deep Dive: From Alert to JVM Memory Leak
An in-depth exploration of a production OOMKilled alert, tracing the issue from JVM memory management to a critical code vulnerability.
Senior Software Engineer at Just Eat Takeaway, where I design event-driven automation so customer issues are resolved without human intervention.
Across 8+ years at Just Eat Takeaway, Grab, Ant Group, and Douban I have scaled backend platforms that support food delivery, mobility, fintech, and consumer communities.
I love end-to-end ownership—pairing resilient architecture with data-informed operations to help cross-functional teams ship measurable impact.
Selected engagements that show how I approach complex operational platforms, automation, and data pipelines in high-scale environments.
Event-driven coordination for Just Eat Takeaway’s customer chatbot. Automated order issue resolution and cut manual interventions by 60% while improving stability for millions of users.
Unified multi-tenant tooling that consolidates 20+ internal services into a single workspace with CQRS-based routing and a shared data model, reducing SLA for issue resolution by a third.
Crowdsourced task management engine at Grab scaling to 300K tasks and 400K+ geospatial points, boosting the accuracy of maps and on-trip experiences across Southeast Asia.
Notes on scaling distributed systems, capturing operational metrics, and the lessons learned while keeping platforms dependable.
5 posts in total
An in-depth exploration of a production OOMKilled alert, tracing the issue from JVM memory management to a critical code vulnerability.
Reflecting on the journey of moving to the Netherlands at 30, embracing new challenges, and pursuing freedom and choice in life.
A guide on initializing Pinia store state using Single-SPA props in a Vue application.