The AI native context-aware semantic cache for LLM apps — Patent Pending - stop paying for the same answer twice