NumPy and Pandas do not benefit from the interpreter changes, and they can even slow down if their compiled wheels do not match your setup.

The async slowdown you see is common. Async helps with I/O, but it does nothing for CPU work. When heavy Pandas or NumPy calls run inside async functions, they block the event loop and make everything feel slower.

The first things worth checking are:

• Reinstall NumPy and Pandas to make sure you have the right builds for 3.11.

• Move CPU-heavy work to a thread pool or process pool instead of running it inside async tasks.

• Profile again under 3.11 because some hot spots shift after an upgrade.

If you want, you can share a small slice of the code or a screenshot of your profiler results, and I can point out the specific parts slowing things down.