Nothing flashy here, just sharing a workflow I've been using myself for some time before applying since most applications end up in major ATS platforms, many of which include AI-assisted parsing, screening, sorting, and/or ranking features.

It's local-first and open-source, meaning no account, hosted DB, or subscription is needed.

The idea is to:

- find out how my PDF CV is parsed, so I can remove my beloved fancy gradient skill bars, etc.

- understand the high-level fit in seconds: no / maybe / worth trying

- see the blockers and what's available to close those gaps, either from added achievements or rough ideas from the model

- get a draft letter with achievements/company context, and a chosen tone with AI patterns cleared [please do not ever send it blindly]

Sadly or not, it won't send 500 applications in a blink of an eye or adapt your CV for you while you look away.

Since it's BYOK, you can use it freely (e.g. using Gemini API), but any provider endpoint you configure may handle the data you send differently.

Curious what you think, and how the flow can be improved.