About Us

Interhuman AI is building the next generation of social intelligence infrastructure—multimodal AI systems that understand not just what humans say, but how they say it. We're developing models that interpret behavioral signals like hesitation, engagement, confusion, and interest across voice, facial expressions, body language, and natural language - in real time. We're looking for an AI Scientist to join our core team and lead the development of models that capture the nuances of human communication in real time.

About the Role

What you'll do

• Design, train, and iterate on speech and language models that extract social and emotional signals from live conversation.
• Own the full model development lifecycle—from data curation and architecture design through training, evaluation, and production deployment.
• Build evaluation frameworks and benchmarks that capture the subtleties of human interaction that standard metrics miss.
• Stay at the frontier of multimodal research and translate relevant advances into our production stack.
• Collaborate closely with engineering to ensure models meet real-time latency and scalability requirements.

Who we're looking for

• PhD in Machine Learning, Computer Science, or a related field with a focus on speech processing and/or NLP.
• Track record of building and shipping models—publications are great, but we care equally about what you've built.
• Strong proficiency in Python and deep experience with PyTorch (or JAX/TensorFlow).
• Familiarity with the current landscape of speech and multimodal models (e.g., Whisper, audio-LLMs, speech encoders, vision-language models).
• You thrive with ambiguity. You can scope your own work, prioritize ruthlessly, and know when to ask for input.
• Clear communicator—you can explain a complex architecture to both engineers and non-technical stakeholders.

What we offer

• Competitive salary and meaningful equity in an early-stage, venture-backed company.