2025-11-05 –, ML+analytics
What if you could make your Pandas code faster, leaner, and cheaper to run?
In this talk, you'll learn how to measure performance, hunt bottlenecks, and optimize your code.
Most of the time, Pandas is fast enough. But when you're working with larger datasets or tighter budgets, "fast enough" isn't enough. In this talk, we'll dive into how to make your Pandas code run faster, and consume less memory. As a bonus, you'll save money on compute costs and reduce experiment times.
We'll kick things off with a quick reality check: when not to optimize (yes, premature optimization is still the root of all evil). Then we'll set performance goals sense, create real-world benchmarks, and learn how to profile your code to find who commits performance crimes.
Finally, we'll explore a bag of tricks and tools to improve performance out of your code. We'll also talk about common gotch that ruin performance.
Previous knowledge expected
Miki has been shipping bugs to production for over 28 years.
He has a passion for teaching, mentoring, and talking about tech for way too long.
Miki contributes to open source, either his own projects, or external ones - including the Go and Python projects.
Miki wrote several technical books, he's a LinkedIn Learning author and an organiser of Go Israel Meetup, GopherCon Israel, and PyData Tel Aviv Conference.
When not geeking out, Miki likes to climb, hike all over the world, read books and annoy his family.