This presentation explores real-world performance and reliability challenges encountered in production systems and how they were solved through practical code optimization and architectural improvements.

Using two concrete case studies, it walks through diagnosing hard-to-explain failures and long-running processes in distributed, event-driven environments (including Kubernetes, AWS S3, and multi-process systems). The talk focuses on how issues were investigated, decisions were made, and solutions were designed-highlighting trade-offs, constraints, and measurable results.

Beyond the technical fixes, the presentation emphasizes a problem-solving mindset: deeply understanding the code, identifying bottlenecks, communicating with stakeholders, and thoughtfully designing solutions before implementing them. Ultimately, it frames code optimization as delivering efficient, resilient solutions that solve real problems while minimizing risk and operational impact.