Velma is Modulate's voice analytics API: instead of transcribing audio and guessing at meaning with a text model, it listens natively - reading emotion, tone and intent from voice conversations and flagging fraud, deepfakes and risk in real time.
Modulate positions it as a cheaper, simpler alternative to chaining a transcription service, an LLM and custom enrichment logic: Velma starts at $0.75 per hour of audio against a claimed $2.50 to $10 per hour for the assembled stack. The company recently announced $25 million in new funding led by Future Ventures to scale its audio-native AI.
