Tavus is an AI video research company building the foundation for human-like, real-time conversational video experiences. Its core technology powers Conversational Video Interfaces (CVI) — AI personas that can see, hear, and respond face-to-face with users in under a second of latency, making them ideal for interactive sales agents, AI tutors, recruiters, healthcare assistants, and customer support. Tavus is best known for its proprietary models: Phoenix-3 for hyper-realistic full-face avatar rendering with natural micro-expressions, Raven-0 for visual perception that gives the AI eyes to understand what users are doing, and Sparrow-0 for ultra-fast turn-taking that enables natural conversational flow. Beyond live conversational agents, Tavus also offers asynchronous video personalization, allowing teams to generate thousands of unique one-to-one videos at scale using a single recorded template — widely used by sales and marketing teams for outbound outreach. The Tavus platform exposes its capabilities through a developer-first API and SDKs, making it straightforward to embed lifelike AI video agents into web, mobile, and embedded products. Backed by top-tier investors including Sequoia and Y Combinator, Tavus has become one of the most cited platforms in the emerging category of conversational AI video.