Last released Nov 7, 2026
Catch spoken PII (card numbers, SSNs, emails said out loud) before voice transcripts reach an LLM. Zero dependencies, Pipecat and LiveKit drop-ins.