Did you (@user1496984) run where python? I think the use case matter before choosing the workflow. I will provide a summarized view inspired from the comments.

Use case = Data # (where # can be analyst, scientist (exhaustively add machine learning engineer, and so on)

• if package manage is brew. Create .env using conda at global level with conda mini-forge base python installation (which shows up w/ where cmd)
• Pros: easily integrates w/ IDEs viz. Zed, Neo-vim, or vim, Cons: not suitable for web-dev

Use case = Web Dev

• create .env at project level using uv, pip or poetry. Personal recommendation: uv
• Pros: You are in control which packages are installed. Cons: If you are new to uv, it could be a bit of a learning curve.

Hope this helps! :)

Debate: Is uv better than conda for Data Science projects?

I am still searching how uv benefits over conda