San Francisco-based David AI, a pioneering audio data research lab backed by Y Combinator, has raised $50 million in a Series B funding round. Led by Meritech Capital and NVIDIA, existing investors such as Alt Capital, First Round Capital, Amplify Partners, and Y Combinator also participated, bringing David AI’s total financing to approximately $80 million since its founding in 2024.
Founded by Tomer Cohen and Ben Wiley, David AI’s ambitious mission is to become the world’s premier source for high-quality and diverse audio datasets. Unlike text or image data, audio data has proven uniquely challenging for AI developers to collect and evaluate rigorously at scale, given its inherent multidimensional nature and real-world complexities.
David AI aims to solve these fundamental problems by building a research-driven lab dedicated solely to the curation, generation, and evaluation of audio datasets that mirror the world’s linguistic and acoustic diversity.
Role of Audio in AI
Audio is widely regarded as the most natural user interface for AI in everyday life. From voice assistants and customer service bots to robotics and wearable devices. David AI’s founders, both alums of Scale AI, contend that audio AI will be transformative in making artificial intelligence accessible and intuitive for billions globally.
“Audio AI has the incredible opportunity to bring AI into the ‘real world,’” the company said in its announcement. “But to live up to this potential, the audio interfaces of the future require exponentially more data and evaluations than are available today.”
That’s where David AI fills the gap, aiming to create the foundational data layer that powers speech models capable of adapting to diverse languages, accents, emotions, cultural contexts, and acoustic conditions. Unlike structured data like code or text, audio and speech are profoundly subjective and contextual.
For example, a phrase that sounds appropriate for a customer support call might be wildly inappropriate as a casual chat with an AI companion. The diversity in tone, pace, emotion, and background noise adds layers of complexity. Multilinguality further compounds this, as accents and dialects cannot be simply translated like text. The deep granularity of audio data requirements makes collecting and validating datasets immensely challenging but essential for building truly generalizable speech AI.
David AI differentiates itself by operating as a dedicated research lab rather than a generic data vendor. Its teams rigorously design, test, and scale datasets with meticulous attention to how data quality influences AI model performance. This research-driven approach allows David AI to serve as a trusted partner for the most advanced AI labs and several leading technology giants, many of which belong to the elite “Mag 7” group dominating AI innovation globally.
The company’s client base spans multiple domains including robotics, generative media, wearables, and automated customer support, underscoring audio AI’s broad applicability. David AI’s datasets empower voice AI models that can authentically understand and participate in real-world conversations and environments, greatly enhancing user experience and AI utility.
Global Collaboration
With its $50 million Series B injection, David AI plans significant expansion across multiple fronts. This includes scaling up its research operations, expanding engineering and product teams, and intensifying collaboration with AI labs and hardware companies worldwide. The fresh capital will also accelerate the development of new evaluation frameworks aimed at measuring real-world model performance more effectively.
The funding round highlights growing investor confidence in audio AI and data-centric approaches, especially as AI companies desperately seek richer, more diverse data to train state-of-the-art models. Participation by NVIDIA hints at synergistic cooperation that could optimize audio AI performance across computing platforms.
David AI has already reached an impressive eight-figure annual revenue run rate, reflecting rapid adoption of its unique datasets and services. The company’s founders remain optimistic about their next growth chapter, envisioning a future where audio AI supports transformational use cases like humanoid robots, advanced wearables, personal AI assistants, and innovative generative media applications.
In an increasingly crowded market for AI data services, David AI faces competition from startups such as Defined.ai, PublicAI, Human Native AI, Ydata, and Syntheticus. However, David AI’s narrow focus on high-fidelity, research-curated audio datasets and its growing relationships with leading AI labs give it a distinctive edge.
