Modulate Raises $25M to Combat Voice AI Fraud and Deepfakes

Overview

Boston-based voice intelligence startup Modulate has secured $25 million in new funding to advance its platform for analyzing and securing voice-based interactions. The company utilizes an array of small, specialized models to provide enterprises with capabilities ranging from transcription and emotional analysis to deepfake detection and policy enforcement. This strategic investment positions Modulate to address growing security concerns as AI-generated audio becomes increasingly indistinguishable from human speech.

The funding round was led by Future Ventures, with participation from Hyperplane and Lakestar. According to data from PitchBook, Modulate had previously raised $41 million at a $170 million valuation, indicating sustained investor confidence in the voice AI sector. This new capital will support the company’s efforts to enhance its detection algorithms and expand its deployment options for regulated industries.

---

Investment and Company Background

Modulate, founded in 2017 by Mike Pappas and Carter Huffman, is pivoting from its original focus to meet the demands of the modern voice AI landscape. Pappas and Huffman met as undergraduate students studying physics at MIT, a background that underpins their approach to modeling complex acoustic signals. While the company initially built tools for voice modulation in gaming, it later shifted toward creating moderation tools for human-driven communications.

"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.

This philosophical shift has defined the company's current trajectory. As generative voice models proliferate, the line between authentic human communication and synthetic audio has blurred. Modulate addresses this by developing technologies that not only transcribe calls but also interpret the intent behind them. The company now operates over 100 individual models, which are broadly categorized into two functional groups:

• Signal Extraction Models: These tools analyze vocal emotion, tone, language patterns, and determine whether a voice is synthetic.
• Analysis and Detection Models: These systems evaluate user intent, identify rule violations, and detect attempts to scam recipients.

The recent $25 million raise aligns with a broader trend among venture capitalists who are backing companies that either improve the human-likeness of AI voices or safeguard against the fraud these technologies enable. Investors are recognizing that as voice interfaces become ubiquitous, the security infrastructure to verify authenticity is just as critical as the interface itself.

---

Technical Approach and Capabilities

A defining characteristic of Modulate’s technology stack is its reliance on smaller, specialized models rather than a single massive foundation model. This architectural choice offers distinct operational advantages in terms of efficiency and scalability. By deploying a suite of narrowly focused AI agents, Modulate reduces its dependency on high-cost specialized hardware and massive compute resources.

This efficiency is becoming a significant economic factor as token costs and energy requirements for large language models continue to rise. The smaller model approach also allows Modulate to rapidly integrate new capabilities. The company can train individual models for specific tasks, add them to its existing portfolio, and use an orchestrator system to call upon the appropriate tool when needed during a live conversation.

The company’s primary use cases revolve around security and quality assurance for enterprise communications:

1. Deepfake Detection: Modulate scans incoming calls to identify synthetic voice markers, alerting call centers to potential scams in real-time.
2. Compliance Monitoring: The system ensures that AI agents interact with customers in accordance with strict regulatory standards, particularly in highly regulated sectors like healthcare and finance.
3. Intent Analysis: Beyond simple sentiment checks, Modulate analyzes whether a customer is genuinely satisfied or merely being polite. This granular data helps organizations understand the root causes of call outcomes.

Huffman noted that traditional sentiment analysis often fails to capture nuance. A customer may use polite language toward an automated agent even when they are significantly dissatisfied. Modulate’s deeper analysis bridges this gap, providing businesses with actionable insights into customer experience. Additionally, the company is leveraging its voice analysis technology to monitor for cyberattacks that originate through voice channels, adding another layer of security to corporate communications.

---

Market Strategy and Future Plans

Modulate’s customer base is diverse, but the company finds its strongest fit in environments where trust and verification are paramount. By positioning itself alongside the voice stack used by enterprises, Modulate acts as a transparent analysis layer that does not disrupt the underlying communication flow. This "sidecar" approach allows businesses to adopt voice AI solutions without sacrificing security oversight.

As more companies deploy AI-powered customer service, the ability to diagnose why a call succeeded or failed has become a key metric for performance. Modulate provides the granular data necessary to optimize these interactions, moving beyond basic metrics to understand human-machine dynamics.

Looking ahead, the startup aims to strengthen its engineering capabilities. The company currently employs between 40 and 45 staff members and plans to hire an additional 10 individuals in the coming months. This expansion is focused specifically on model building and refining the analytical engine.

Privacy remains a top priority in Modulate’s development roadmap. The company is actively working on enhancing its on-premises and on-device deployment capabilities. This move is designed to cater to clients with strict data sovereignty requirements, allowing sensitive voice data to be processed locally rather than in the cloud. By combining advanced intent analysis with robust security features, Modulate is positioning itself as a critical infrastructure provider in the evolving ecosystem of voice-based artificial intelligence.