Boston-based voice intelligence startup Modulate has raised $25M in new funding for a platform that combines transcription, emotional analysis, deepfake and AI-music detection, and policy enforcement for voice agents in regulated industries. The round was led by Future Ventures, with participation from Hyperplane and Lakestar. PitchBook data cited by TechCrunch indicates the startup had raised $41 million at a $170 million valuation before this round.

Modulate was founded in 2017 by Mike Pappas and Carter Huffman, who met as MIT physics undergrads. The company initially worked on voice modulation for gaming, then shifted to a voice-based moderation tool. With the rise of voice AI models, Modulate now focuses on detecting different kinds of AI audio generation and analyzing the intent behind a person’s words. Huffman told TechCrunch that many companies do transcription, but few capture the full nuance of a conversation, which matters when humans talk to each other.

The company runs more than 100 models in two main categories. Signal extraction models understand vocal emotion, tone, language, and whether a voice is synthetic. Analysis and detection models assess intent, such as what a customer is trying to say, whether a caller is violating rules, or whether they are attempting to scam the receiver. Because the models are small, Huffman said, Modulate does not need specialized hardware or large amounts of compute, which could become important as token costs rise. The smaller-model approach also makes it easier to train new capabilities, add them to the collection, and have an orchestrator call them when needed.

Modulate has a varied customer base but specializes in deepfake detection and alerting organizations like call centers to possible scams. It also monitors how AI agents respond to customers to measure call quality and checks that AI follows compliance rules in regulated areas. As a result, Modulate often sits alongside a company’s existing voice stack purely to analyze calls. Huffman noted that simple emotion analysis can be misleading: customers may remain polite to AI agents while still being dissatisfied, so granular intent and response data matter.

The company says its technology is also used to monitor cyberattacks conducted through voice calls. Modulate currently has 40-45 employees and plans to add about 10 more in the coming months to strengthen model building. It is also working to improve on-premises and on-device deployment for increased privacy.