About the role

Your mission will be to transform Neon’s unique data streams into the gold standard for AI training and evaluation. As a Senior Speech AI Researcher, you will bridge the gap between our 500,000+ mobile users and the world’s leading AI labs, by directly interfacing and collaborating with your peers in these labs. You will define our research agenda, establishing Neon as a premier authority in audio data research and evaluation science.

You’ll lead the development of AI assisted evaluation pipelines to characterize and validate the data our platform generates, ensuring it solves the "data bottleneck" for frontier AI models. This is a high-impact, foundational role where you will be the face of Neon within the global AI research community, translating theoretical data needs into tactical product execution.

We're a small, early-stage team moving fast. Expect roughly 60 hours per week, with occasional weekend work around launches and deadlines. We're upfront about this because we'd rather find people who are energized by that pace than surprise anyone after they join.

You have…

• Authorization to work in the US.
• Track record of impactful AI/ML research, specifically focused on data evaluation, oversight, and generative AI.
• Significant audio and multimodal expertise, able to understand and clearly articulate the latest developments in the space. Speech specific expertise is required.
• Deep engagement with the AI research community; you speak the language of frontier labs and understand the current trends in model training and evaluation.
• A "Founder’s Mentality." You are comfortable with ambiguity and have a proven ability to translate complex research insights into production-ready data products.
• Exceptional communication skills, capable of building relationships with top-tier researchers while clearly articulating Neon’s value proposition to partners.
• Experience with Python and PyTorch, designing and conducting rigorous ML experiments and evaluations.

Bonus points

• Published authorship in top-tier AI/ML venues such as NeurIPS, ICML, ICLR, or CVPR.
• Strong record of publications in data evaluations, especially for audio/speech, or multimodal data.
• Experience building technical communities or academic advisory boards from the ground up.
• Specific expertise in evaluation science, including AI-assisted evaluation (RLAIF) or designing benchmarks for frontier capabilities.
• An active presence on X or other research channels, demonstrating thought leadership in the AI data conversation.

Compensation

Competitive salary & equity.