Hello everyone, I’m reaching out to inquire if anyone here is involved in the field of LLM guardrails. Specifically, I’m interested in exploring ways to enhance the security of LLMs by preventing harmful prompts and outputs. Additionally, I’m keen on discussing strategies to safeguard sensitive information within this context. Looking forward to potentially collaborating and sharing ideas.

have you looked into https://github.com/guardrails-ai/guardrails

Yes I checked, Guardrails.ai and NeMO from Nvidia