Oruk is a speech AI lab building systems for real-time conversation understanding. We focus on the acoustic signals that shape human meaning: timing, tone, hesitation, uncertainty, affect, and intent. That layer is where speech carries true context and emotion.

Our work is built for infrastructure teams and operators, bringing deeper understanding into real products. From voice agents and customer support to healthcare triage, ordering systems, games, avatars, and call analytics.

Oruk is built by a research team with roots at Stanford and Berkeley. We are exploring the next layer of speech behavior and sharing our research notes, benchmark updates, and milestones openly as we build.

Stay up to date with our research blog: oruk.ai/research

Oruk

Org chart

Teams

Offices