ElevenLabs Releases Agent Orchestration Engine
ElevenLabs has released ElevenAgent, an orchestration engine for building production-grade multi-agent systems. The architecture emphasizes persistent, auditable agent state, dynamic skill registration, and explicit agent specialization—key patterns for building compliant and debuggable workflows in regulated industries like insurance.
The move comes as ElevenLabs solidifies its market position, having raised a $500 million Series D at an $11 billion valuation, led by Sequoia Capital. The company, founded in 2022, has now raised a total of $781 million and closed 2025 with over $330 million in annual recurring revenue, signaling significant enterprise adoption. ElevenAgent's architecture is engineered for real-time conversation, adding less than 100ms of overhead by using a multi-model pipeline. This approach combines proprietary models for tasks like turn-taking with frontier LLMs from providers like Google and OpenAI, optimizing for the trade-offs between response speed, intelligence, and cost. This reflects a broader industry shift from simple prompt chains to stateful, graph-based agent workflows, a design pattern better suited for production reliability. Leading open-source frameworks like Microsoft's AutoGen and LangGraph are enabling developers to treat multi-agent system design as a distributed systems problem, focusing on orchestration, state management, and observability rather than just prompt engineering. In insurtech, these agentic architectures function as an intelligence layer orchestrating workflows across legacy policy and claims systems. Commercial P
