Peter Chan January 18, 2025 10:38
Eleven Labs has successfully deployed voice agents that resolve over 80% of daily user inquiries, enhancing customer support through AI-driven solutions.
Eleven Labs deployed a voice agent designed to efficiently handle user inquiries regarding documents, and reportedly achieved a resolution rate of over 80%. Voice agents handle approximately 200 calls each day and have great success in responding to user queries.
Performance and evaluation
Voice agents powered by large-scale language models (LLMs) are recognized for their ability to effectively resolve or redirect inquiries. Human validation of 150 conversations revealed 81% agreement between LLM and human raters for successfully resolved queries. Agents also showed 83% agreement in maintaining knowledge-based compliance.
Additionally, 89% of relevant support questions are answered or correctly redirected by the documentation agent, demonstrating their ability to manage simple queries.
strengths and limitations
Strengths
LLM-powered agents are better at solving specific questions that match available documentation. It effectively directs users to relevant pages, provides initial guidance on complex queries, and has proven beneficial for questions such as API endpoints, language support, and integrated queries.
To optimize performance, Eleven Labs recommends targeting users with clear questions and utilizing redirects for more complex inquiries to increase the efficiency of the support process.
Restrictions
Despite its strengths, the agent faces challenges with vague inquiries and account-related inquiries that require deeper investigation. Since voice media is not well-suited for sharing code or handling complex technical issues, Eleven Labs suggests redirecting users to documentation and support channels for such inquiries.
Development and configuration
The voice agent is configured with system prompts to guide your responses and keep you focused on elevenLabs products. A comprehensive knowledge base with condensed versions of all documents helps LLMs provide accurate answers.
Three key tools are integrated into the agent’s functionality: redirection to external URLs, email support, and documentation, providing a versatile route for user inquiries. Agent rating tools evaluate conversations against predefined criteria to ensure continuous improvement and reliability.
continuous improvement
Eleven Labs recognizes the limitations of LLM in solving all kinds of queries, especially in a rapidly evolving startup environment. But the company emphasizes the benefits of automation, which frees up teams to focus on complex challenges as the community expands the possibilities of AI audio technology.
ElevenLab’s Conversational AI-powered agent serves as an effective tool for resolving product and support questions, and is continuously improved through automated and manual monitoring to improve the user support experience, the company says. It reflects our efforts.
Image source: Shutterstock
Source link
