Rust is something you may already be using every day: it powers uv, ruff, and polars. After pydantic v2 rewrote its validation core in Rust, it became 17 times faster. What’s even more noteworthy is that the official CPython “Rust for CPython” working group has confirmed its target is Python 3.16. They plan to experimentally rewrite the zlib module using “zlib-rs”, with the related PEP expected to be submitted for community discussion this summer. If this lands, every “pip install” decompression you perform will become faster.
Rust is not Python’s rival, but its partner: Python handles orchestration, while Rust delivers speed and safety. This lightning talk is based on my real experience as a Python developer self-teaching Rust. In 10 minutes, I will share three things:
- Why it’s worth learning: even if you never write production-grade Rust, the habits of ownership, explicit error handling, and exhaustive matching will transfer back to make your Python code more rigorous (in the free-threading era, this kind of concurrency safety intuition is especially valuable).
- Practical implementation: rewriting a real CPU-intensive hot function as a Rust extension using PyO3 + maturin. On the Python side, only one import line needs to change. Tests pass unchanged. I will show before-and-after performance and memory measurements, along with a decision checklist for “which code is worth rewriting and which is not.”
- One-month learning roadmap: where the mindset shift from EAFP to LBYL is difficult, how to overcome the borrow checker hurdle, and a four-week study plan based on Microsoft’s open-source course “Rust for Python Programmers” and the official Rust Book.
No Rust background is required from the audience. You will leave with an immediately actionable learning roadmap, a set of step-by-step instructions for migrating hot functions, and a decision framework for judging “whether Rust should be used or not.”