Artificial Intelligence
LLM Guardrails in Production: What Actually Works
Prompt injection, hallucination detection, output validation. We shipped six production LLM systems in the past year. Here is what actually works and what only sounds good in blog posts.
Linh PhamHead of AI/ML Engineering2024-12-109 min read
Scroll to explore
Prompt injection, hallucination detection, output validation. We shipped six production LLM systems in the past year. Here is what actually works and what only sounds good in blog posts.