is that intentional?

I think Frank might've done it to keep things simple for now. We'll need to benchmark things to find where most of the time is being spent. I think we can get to it after Python is implemented.

ah gotcha makes sense 👍

yeah, I thought about keeping a process around and re-using it for subsequent requests. But then I realized I would have to keep more than 1 process around to handle concurrent requests. In the end, I might have to end up managing a process pool. Which is probably a good idea down the road.