PromptGuard - AI Prompt Security is an AI-powered, drop-in "firewall" for LLMs that inspects and sanitizes prompts and surrounding context in real time—using heuristics, ML classifiers and LLM-based detectors—to block prompt injection, redact PII, and prevent data leaks before requests reach your model. It adds minimal latency (<40ms), provides tunable policies, logs and analytics, supports major providers (OpenAI, Anthropic, Google/Gemini, Groq, Azure), and is aimed at product, security and enterprise teams that need production-grade prompt governance and compliance.