# Efficient Document Support Voice Agent by ElevenLabs
**Who:** ElevenLabs
**What:** Successfully implemented a voice agent resolving over 80% of user inquiries daily
**When:** Jan 18, 2025
**Where:** Global
**Why:** Enhancing customer support through AI-driven solutions
**How:** By introducing a voice agent designed to efficiently handle user inquiries related to its documentation
**ElevenLabs** has revolutionized customer support with its latest implementation of a cutting-edge voice agent, effectively addressing over 80% of user inquiries daily. This innovative solution marks a significant milestone in enhancing customer experiences through AI-driven technologies.
## Performance and Evaluation
The voice agent, powered by a large language model (LLM), has undergone rigorous evaluation to determine its effectiveness in resolving user inquiries. In a study involving 150 conversations, the LLM and human evaluators reached an impressive 81% agreement rate on successfully resolved inquiries. Additionally, the agent demonstrated an 83% agreement on maintaining adherence to the knowledge base, showcasing its reliability and accuracy.
Moreover, the voice agent excelled in addressing 89% of relevant support questions, either providing answers or correctly redirecting users to appropriate resources. This remarkable performance underscores the agent’s capability in managing a wide range of user inquiries efficiently.
## Strengths and Limitations
### Strengths
The LLM-powered voice agent shines in providing precise answers to specific questions that align with the available documentation. Its ability to direct users to relevant pages and offer initial guidance on complex queries, such as API endpoints and integration issues, proves highly beneficial for users seeking technical support.
To enhance its performance further, ElevenLabs recommends targeting users with clear questions and utilizing redirects for more complex inquiries, streamlining the support process and maximizing efficiency.
### Limitations
While the voice agent excels in many areas, it faces challenges with vague or account-related inquiries that require in-depth investigation. Additionally, handling code-sharing or complex technical issues via voice may not be ideal, prompting ElevenLabs to suggest redirecting users to documentation or support channels for such queries.
## Development and Configuration
The voice agent is meticulously configured with a system prompt to guide its responses, ensuring a focused approach towards addressing user inquiries related to ElevenLabs’ products. Backed by a comprehensive knowledge base that summarizes all documentation, the LLM delivers accurate and insightful answers to users.
Three primary tools integrated into the agent’s functionality—redirecting to external URLs, email support, and documentation—offer users versatile pathways to seek assistance. The agent’s evaluation tooling continuously assesses conversations against predefined criteria, facilitating ongoing improvement and reliability.
## Continuous Improvement
Acknowledging the evolving nature of user queries and advancements in AI technology, ElevenLabs remains committed to enhancing its voice agent’s capabilities. While recognizing the limitations of LLMs in handling all query types, the company emphasizes the value of automation in streamlining user support processes.
Driven by ElevenLabs Conversational AI, the voice agent represents a powerful tool for navigating product and support inquiries, continually refined through automated and manual monitoring. This dedication to enhancing user support experiences underscores ElevenLabs’ commitment to innovation and customer satisfaction.
In a world where technology is rapidly evolving, ElevenLabs stands at the forefront of revolutionizing customer support through its efficient and intelligent voice agent. As the company continues to push boundaries and explore new frontiers in AI-driven solutions, users can expect unparalleled support and seamless experiences in their interactions with ElevenLabs’ products.
*Image source: Shutterstock*
**Keywords:** **voice agent**, **ai**, **customer support**
