Ultimately the management should care about what brings in the most revenue. A feature that brings low revenue but has no fire is something no-one cares about. If you have a ton of revenue and no bugs, you likely can get recognition.
> This has a great SQL-ish API. Python is similar but starts to be a little clunky since it requires you to think about indices:
groupby has an as_index parameter for this very purpose
> Deducting the discount
You focus on doing the subtraction during the group by. Is there any good reason for this? You could either do it as a step before, or after summing up both columns. Putting too many things into one command is not good practice yet you benchmark the language based on how easy it is to do said bad practice
> The value-add that Microsoft brings to Github Copilot is near zero compared to directly buying from Anthropic or OpenAI
Over here in the EU, we need to store sensitive data in an EU server. Anthropic only offers US-hosted version of their models, while G-cloud and Azure has EU based servers.
I don't think the no moat approach makes sense. In a world where more an more content and interaction is done with and via LLMs, the data of your users chatting with your LLM is a super valuable dataset.