Build conversational AI assistants now with ElevenLabs

ElevenLabs now offers voice cloning service with artificial intelligence and text-to-speech API. This company has recently introduced the ability to build conversational bots with artificial intelligence.

According to TechCrunch, the startup announced that users can now build a conversational agent (Conversational Agent) on the ElevenLabs developer platform, which includes adjustable variables such as tone of voice and length of responses.

Until now, ElevenLabs focused on providing various voices and artificial intelligence tools for text-to-speech services. According to Sam Sklar, the company’s director of growth, many customers used this feature to create conversational agents. But the main challenge was integrating the knowledge base and managing customer interruptions. For this reason, the company decided to build a complete production line for processing conversational bots.

New features for building a conversational assistant

Users can create a conversational assistant by logging into their ElevenLabs account. By selecting a template or creating a new project, they can specify settings such as the agent’s primary language, initial message, and system prompts to determine the assistant’s personality. Developers should also set a large language model such as Gemini or Claude, the level of response creativity and the limit of token usage. Other features of this service are as follows:

• Settings such as volume, response delay, persistence, authentication criteria, and maximum conversation length.
• Add a custom knowledge base containing files, links or text blocks to support the conversational bot.
• Integrating custom language model with bot.

Companies can define criteria for collecting specific information from customers (such as name and email) as well as evaluating the success or failure of a conversation in natural language.

Development of speech skillsIn artificial intelligence assistant

ElevenLabs uses a dedicated text-to-speech product line, but for its new product, it has also developed speech-to-text capabilities. Although the API does not currently offer speech-to-text conversion independently, it may do so in the future, competing with services from companies such as Amazon, Microsoft, and Google, as well as specialized APIs such as Whisper (OpenAI), AssemblyAI, Deepgram, Speechmatics, and Gladia. slow

The company, which is looking to raise new capital at a valuation of more than $3 billion, competes with other startups such as Vapi and Retell, which make conversational assistants. It will also compete with OpenAI’s real-time conversation API. However, ElevenLabs believes that customization capabilities and the ability to change models will be its competitive advantage over OpenAI.