MODULATE , a frontier AI company focused on analysing voice, has raised an additional $25 million from Future Ventures, with participation from Hyperplane and Lakestar. The investment brings its total funding to $60 million and will support expansion of its team and infrastructure, as well as further development of its models, APIs, software development kits, integrations and deployment options.
The company’s Velma platform uses a proprietary Ensemble Listening Model architecture and more than 100 specialised models to assess emotion, tone, intent, synthetic speech and conversational behaviour in real time. Modulate says it analyses more than 10 million hours of audio each month and has processed over 600 million hours in total.
It also claims its transcription and deepfake-detection models ranked first on public benchmarks including Hugging Face, while Velma is twice as accurate as traditional large language models and produces seven times fewer false positives. These are company claims; the article provides no independent verification or evidence of confirmed attacks detected by the platform.
Modulate says its technology is being applied to protect healthcare organisations from deepfake attacks, identify extremism, harassment and child-grooming conversations on social platforms, improve the emotional responses of voice agents, monitor agent performance and protect high-risk workers from voice identification through masking. Chief executive Carter Huffman said audio analysis is necessary because transcripts cannot capture all the signals in a conversation.