ElevenLabs Unveils Agent Transfer Feature to Share Data Between AI Agents

ElevenLabs introduced a new enterprise-focused feature on Monday that will allow artificial intelligence (AI) agents to communicate with each other. Dubbed Agent Transfer, the feature transfers a conversation from one agent to another when specific conditions are met. The first agent also shares the previous conversation data with the onboarding agent to help it contextualise the information and carry forward the conversation. The AI audio-focused company says this feature will allow businesses to build parallel AI agents with different specialisations.

ElevenLabs Introduces a New Agentic Feature

In a post on X (formerly known as Twitter), the AI firm announced the release of the new agentic feature. Agent Transfer is part of Conversational AI and allows two AI agents to communicate and share conversations. It is currently available to enterprises, but the company has not mentioned if the capability will be priced separately or will be part of the existing plans. ElevenLabs has also detailed how developers can set this capability via the Conversational AI in a support page.

As businesses increasingly adopt AI agents into their internal and external workflows, there is a need to avoid data silos, so that the information does not remain limited to one business function while being unavailable to the rest. While different AI giants are coming up with different solutions to create centralised data hubs, ElevenLabs is trying to get agents to talk to each other to share data.

How it will work is interesting. Imagine the first AI agent is having a conversation that leads to an escalation or a scenario where its knowledge base is not able to assist the human user any further. Typically, AI agents hand over the conversation to a human agent who then takes the conversation forward, or they register the request or complaint for a human agent to review later.

With Agent Transfer, the AI agent will be able to transfer the conversation to another agent, which might have the knowledge to better assist the user. The first chatbot will also transfer the conversation history, allowing the onboarding agent to gain full context of the conversation so far.

One key area where such a system would benefit businesses is the customer support department. Enterprises could build specialised AI agents capable of handling general queries, technical queries, complaints, and so on. By running these AI systems in parallel, businesses can further optimise the operation, the company claimed.