An innovative step forward in the realm of artificial intelligence has been made by Modulate, an AI pioneer, as they secure $25 million in funding from Future Ventures. This round, which also saw contributions from Hyperplane and Lakestar, elevates Modulate’s total financial backing to $60 million. The company is renowned for developing audio-native AI models that proficiently distinguish between genuine and synthetic human speech.
Advancements in Audio AI Technology
At the core of Modulate’s breakthrough is their proprietary Ensemble Listening Model (ELM) architecture. This technology powers the Velma platform, which integrates over 100 specialized AI models. These models are designed to comprehend and identify elements such as emotion, tone, intent, and synthetic speech, thereby enhancing conversational behavior analysis.
Currently, Modulate’s technology processes over 10 million hours of audio monthly, amassing a total of 600 million hours analyzed to date. The company’s transcription and deepfake detection capabilities have achieved top rankings on public benchmarks, including those hosted by Hugging Face.
Real-Time Applications and Impact
Modulate’s focus remains on analyzing rather than generating voice, claiming their Velma platform offers twice the accuracy of conventional large language models (LLMs) with significantly fewer false positives. Operating in real time, Velma not only comprehends ongoing conversations but also allows for immediate intervention when necessary.
The applications of Modulate’s technology are vast and impactful. They include safeguarding healthcare systems from deepfake threats, bolstering emotional and empathetic responses in voice AI, mitigating online extremism and harassment, and enhancing protective measures for agents in high-risk situations through advanced voice masking.
Future Outlook and Strategic Developments
CEO Carter Huffman emphasizes the transformative role of voice as an AI interface. “With audio-native AI, we aim to protect organizations from deepfake attacks and enable voice agents to respond more empathetically and accurately,” he explains. The new funding will be instrumental in expanding Modulate’s team and technological infrastructure, further developing their models, APIs, SDKs, and deployment options.
The importance of AI in battling fraudulent deepfakes and abusive language misuse is undeniable. Modulate aims to simplify the process for developers by providing a robust audio intelligence layer, thereby allowing them to focus on creating innovative voice experiences. Huffman adds, “We have the technology proven at scale, and this investment enables us to grow rapidly to meet the increasing demand.”
As the landscape of audio-native AI continues to grow, Modulate stands at the forefront, ready to drive the next wave of innovation in deepfake detection and voice application development.
