Modulate raises $25 million for audio-native AI models

Boston-based AI startup Modulate has raised $25 million in a funding round led by Future Ventures, with participation from Hyperplane and Lakestar. This brings the company’s total funding to $60 million.

Modulate develops audio-native AI models designed to analyze raw audio signals rather than relying solely on text transcripts. Its flagship Velma platform identifies emotion, tone, intent, emphasis, and synthetic speech. This technology is used for fraud and deepfake detection, AI agent supervision, customer experience, and trust and safety applications.

The company’s Ensemble Listening Model architecture utilizes over 100 specialized audio models, which Modulate claims is up to 1,000 times more efficient than using a single large model. Currently, the company processes more than 10 million hours of audio monthly, with a lifetime total exceeding 600 million hours. Modulate also reported that its transcription and deepfake detection technologies have achieved top rankings on Hugging Face benchmarks.