Urgent.News

What's breaking now, across thousands of outlets.

AI

20 minutes with the CEO of ElevenLabs, now reportedly valued at $22B

ElevenLabs powers the AI voice on the other end of a lot of customer service calls, and its CEO told me this week that businesses should probably tell you that — at least until getting a machine is what everyone expects anyway.

ElevenLabs, a company building the voice layer for AI, has gained significant attention after being reportedly valued at $22 billion by its investors. Despite being just four years old, the firm has an annual recurring revenue of $600 million. ElevenLabs' technology powers customer service for companies like Klarna, Deutsche Telekom, Cisco, and Adobe, as well as governmental entities.

The company faces stiff competition from other players, including Decagon, which has incorporated ElevenLabs' technology into its own voice product. However, ElevenLabs' co-founder and CEO, Mati Staniszewski, remains undeterred and sees future growth opportunities by passing the Turing test for conversational AI. He believes that combining intelligence with emotional intelligence will set ElevenLabs apart and usher in a new era of customer interactions.

Staniszewski also discusses the blurry lines between model, platform, and application companies in today's AI landscape, highlighting ElevenLabs' approach to catering to various industries' requirements, from open-source models for less sensitive inquiries to closed-source or fine-tuned models for more critical scenarios. He emphasizes the importance of disclosing when customers are interacting with an AI agent, stating that it will eventually become the norm as the technology becomes more prevalent.

Written by urgent.news from TechCrunch's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at techcrunch.com →

More in AI

Running a Jev-Style Decision Model on One TPU v6e: What Fits, What It Costs, and What Changes From a GPU

This article provides a step by step guide to running a Jev-style decision model on one Google Cloud TPU v6e chip with Gemma 4 and vLLM, and compares it with the same read on one NVIDIA L4 GPU.

  • Jev-style decision model runs on Google Cloud TPU v6e using Gemma 4 and vLLM
  • TPU serves 12B models in E2B, E4B, and 12B configurations in bf16 precision
  • GPU can handle same models in bf16 precision, but 26B-A4B and 31B exceed TPU memory

More from Thursday 24 September →