Modulate raises $25M for its voice models and analysis suite

Boston-based voice intelligence startup Modulate has raised $25 million in new funding. The company uses an array of small models to give enterprises transcription, emotional analysis, deepfake and AI music detection, and policy enforcement for voice agents in regulated industries.

Who backed the round

Modulate's funding was led by Future Ventures, with participation from Hyperplane and Lakestar. Data from PitchBook indicates the startup had raised $41 million at a $170 million valuation before this round.

Mike Pappas and Carter Huffman founded the company in 2017 after meeting as MIT physics undergraduates. In its early days Modulate worked on voice modulation for gaming, then moved toward a voice-based moderation tool. With the arrival of voice AI models, it now focuses on detecting AI-generated audio and reading the intent behind a person's words.

More than 100 small models

Modulate runs more than 100 models divided into two broad groups. Signal extraction models read vocal emotion, tone, language and synthetic voice. Analysis and detection models look at intent: what a customer is trying to say, whether a caller is violating rules, or whether they are trying to scam the recipient.

Huffman told TechCrunch that running smaller models means the company does not need specialized hardware or large amounts of compute, an advantage as token bills rise.

"Our insight into the voice AI space is that a lot of folks are doing transcription, but there's not really any capability out there that gets the full nuance and full understanding of a conversation, which is so important when you're talking to another human being," Huffman said.

What enterprises use it for

The customer base is varied, but Modulate specializes in deepfake detection and in alerting call centers to possible scams. It also monitors how AI agents respond to customers to assess call quality, and checks that AI follows compliance rules in regulated areas. In many deployments it sits beside the voice stack a company already uses and analyzes calls rather than replacing it.

As more enterprises adopt AI-powered customer service, understanding why a call succeeded or failed becomes more important. Huffman said sentiment labels alone can mislead, because customers are often polite even to bots: they may not sound angry while being deeply dissatisfied. Modulate says its technology is also used to monitor cyberattacks carried out through voice calls.

The startup has 40-45 employees and plans to add about 10 more in the coming months to strengthen model building. It is also working on on-premises and on-device deployment for greater privacy.