ElevenLabs : How Its Enterprise Clients Are Putting Its AI to Work

ElevenLabs : How Its Enterprise Clients Are Putting Its AI to Work
3 min read

ElevenLabs highlighted several enterprise clients using its voice agents and voice models at its ElevenLabs Summit NYC in New York. According to the company, a useful AI must be able to speak like a person, fully understand context, and collaborate with humans by assigning them tasks and handing off when it reaches its limits.

Ashley Kramer, Chief Revenue Officer at ElevenLabs, believes large language models (LLMs, AI systems trained on vast amounts of text) are “incredibly intelligent.” “Now, it’s about how to keep up with demand for capacity, how to make tokens (the units of text processed by the model) more cost-effective, and how to ensure the models serve people at the level of intelligence they need,” she said.

She says the company’s latest v4 models are more expressive and context-aware, capable of retaining information from one call to the next for the same customer. “We are going to enter an era where people will prefer to speak to an AI agent in some cases, because you can get things resolved faster,” she added.

Enterprises now account for over 55% of ElevenLabs’ revenue. According to the company, its technology is part of the daily operations of five of the top ten tech companies, five of the top ten insurers, and four of the top ten telecom operators.

Banking, Services, and Media

Customers Bank, a Pennsylvania-based bank, uses several ElevenLabs agents. Among them is a 24/7 voice and messaging agent for account, card, and transfer inquiries via phone, web, and mobile. Carlos Coto, the bank’s Chief Strategy & Innovation Officer, summarized: “You need to offer services where customers want to interact: full self-service for those who don’t want to talk to a human, and people assisted by agents for those who value the exchange.”

Banco Inter, a Brazilian digital bank with a presence in the United States and Argentina, has 46 million customers. It handles over 2 million customer service interactions per month in Portuguese, English, and Spanish using AI. Daniela Mineiro, the bank’s Customer Experience Manager, states that satisfaction rates are higher when the response is provided by AI than when the request is transferred to a human.

At Yahoo, the Yahoo Scout answers tool received voice capabilities via ElevenAgents, enabling real-time conversations in addition to text. According to Eric Feng, SVP and GM of the Yahoo Research Group, the company could not have launched this product as quickly without this technology. Yahoo Scout has now exited beta after providing over 3 billion answers.

Westgate Resorts, which hosts over two million owners and guests annually across nearly 60 resorts in North America, developed a voice agent named “Sarah” with ElevenAgents. Trained on 8,000 transcripts from its actual support calls, she handles common front desk requests. She was launched at its Orlando resort, which has 2,100 rooms, and her deployment is set to expand to two more resorts.

Media, Healthcare, and Telecoms

FOX Sports integrated a conversational digital persona of commentator Colin Cowherd into its app, offering interactive real-time commentary. Dustin Myers, EVP of Production Operations and Technology at the network, emphasizes that a broadcast must “resonate with that market,” not just be translated.

Clover Health, a healthcare technology and Medicare insurance company offering health plans and clinical software to seniors, is interested in the linguistic aspect of voice. Cameron Pringle, Head of Strategy, Office of the Chairman, states that “overnight, we can activate 35 languages.”

Hasbro had the voices of its iconic characters recreated with ElevenLabs as part of its AI studio, Sixth Wall. Roberta Thomson, CEO of Sixth Wall, speaks of “Behavioral Licensing”: the license no longer just defines where a character appears, but also how it behaves in a conversation with a customer.

Finally, KPN, the leading telecom operator in the Netherlands, launched Mobiel AI with ElevenLabs. According to the two companies, fine-tuning the model on Dutch data increased identity verification accuracy from about 50% to 90%.

S’abonner
Notification pour
guest

0 Commentaires
Le plus ancien
Le plus récent Le plus populaire
0
Nous aimerions avoir votre avis, veuillez laisser un commentaire.x