#210: Codeflash and continuous Python performance with Saurabh Misra

Speed isn’t just a nice-to-have - it affects user experience, cloud costs, and how fast teams can move. In this episode, we chat with Saurabh Misra about making Python performance a continuous habit rather than a last-minute clean-up. He introduces Codeflash, a tool that profiles real code paths, explores optimisation options with LLMs, and only suggests changes that preserve behaviour and deliver measurable speedups. We delve into how this works, from tracing and line-level profiling to coverage-guided inputs and concolic testing. Saurabh shares real examples, including smarter NumPy usage, avoiding unnecessary global sorts, and using Numba to speed up numeric hotspots. We also talk about fitting performance checks into everyday workflows via the CLI, VS Code, and GitHub Actions. The big takeaway: performance doesn’t have to slow teams down — with the right tooling, it can be part of shipping well from day one. Connect with Saurabh at https://www.linkedin.com/in/saurabh-misra/ and find out more about Codeflash via the website https://www.codeflash.ai/. ___ 💡🧑‍💻 Want to become a more focused, motivated, and effective Python developer? The Pybites Developer Mindset (PDM) Program helps you build strong habits, deepen your skills, and make real progress in just six weeks. Join us for mentorship, accountability, and a supportive community that keeps you moving forward. Start your PDM journey now! 🌟✅ https://pybit.es/catalogue/the-pdm-program/ ___ If you found this podcast helpful, please consider following us! Start Here with Pybites: https://pybit.es