nandhakishorm/laya ? reverse-engineered prompt

Reverse engineered prompt

Build me a Python library that can read a piece of text and quickly make typed decisions from it in one pass. I want it to handle things like choosing a category, giving a score, and answering yes or no questions, with a simple router that sends English text to one model and other languages to a multilingual model automatically.

It should work on short support tickets, long documents, and at least 100 languages. Please include a clean Python API, a small command line tool, and a basic server option so I can call it from another app. Make the results easy to use in code, with clear labels and confidence style outputs.

Also add a few examples, some tests, and simple docs that show how to use it for things like support triage or churn risk detection. If current model or serving docs matter, look them up online and follow the latest approach.

Are you gonna build this?

make sure you review the code using arcumet

Try freeSponsored — opens Arcumet in a new tab