Putting the Pro in Profiles
Continuous Profiling is the next frontier in observability, promising deep insights into system performance with minimal overhead. Yet for many developers, the thought of flamegraphs, stacks, and CPU cycles can feel like black magic. In this talk, we’ll break down profiling into something not just accessible, but empowering. You’ll discover why profiles are a missing pillar of observability, how continuous profiling changes the game compared to traditional tools, and why flamegraphs — despite their intimidating looks — are actually intuitive and powerful once demystified. Along the way, we’ll explore practical examples where profiling turned sluggish systems into high-performers and revealed optimization opportunities worth millions. Key takeaways: Understand how to use flamegraphs to spot and fix performance bottlenecks fast Learn practical techniques to optimize CPU, memory, and network resource usage See how continuous profiling transforms system observability and cost efficiency Gain confidence to apply profiling insights in production environments at scale By the end of this talk, you’ll be able to decode flamegraphs with confidence, uncover hidden performance issues across your stack, and apply continuous profiling to drive real improvements.
Speaker
-
Ismail SimsekGrafana LabsIsmail is a Staff Software Engineer in the OSS Bigtent squad, working on open-source observability tools from Munich, Germany. He focuses on building reliable systems, improving developer workflows, and making complex problems easier to solve. Outside of work, he enjoys cooking with his wife, getting lost in good books, and creating imaginative role-playing adventures for his daughter.