ElevenLabs CEO: Voice will be the fundamental interface for tech

Author

Ryan Howard
July 07, 2025

🔊 Soundcheck

  • ElevenLabs CEO outlines why voice is the next interface

  • Intron's voice AI expands beyond healthcare

  • Free AI chat in German, no login required

🔥 Hot Mic

Big moves, deep dives, and standout stories.

In a wide-ranging interview on Sequoia’s Training Data podcast, ElevenLabs co-founder and CEO Mati Staniszewski shared why the company is doubling down on voice while others go multimodal, and how that focus is paying off.

Founded by two high school friends from Poland, ElevenLabs has quickly become a leader in AI voice generation. While major labs chased broader models, ElevenLabs specialized in audio, applying transformer and diffusion architectures to build context-aware, emotionally rich text-to-speech systems. This technical edge, combined with remote-first hiring and fast product iteration, allowed them to outperform larger competitors in expressive, real-time voice synthesis.

But the real unlock? Prosumers. From viral videos like Harry Potter by Balenciaga to audiobook authors hacking their way to professional narration, the company's bottom-up approach surfaced new use cases that later drove enterprise adoption. ElevenLabs’ voice is now found everywhere from YouTube channels to Epic Games’ Darth Vader in Fortnite.

Staniszewski outlined their vision of voice as the default interface for technology: personal tutors, real-time translation, and voice agents that can automate everything from customer support to chess instruction. He also emphasized the importance of deep integration, building developer tools, safety layers, labeling workflows, and real-time translation pipelines around the core model to support serious use cases.

Key points:
• ElevenLabs stayed audio-focused while others chased multimodal models
• Prosumer adoption led to viral growth and surfaced enterprise use cases
• Built an infrastructure layer: voice coaching, labeling, dev tools, and safety controls
• Sees voice as the natural interface for education, agents, and global communication
• Predicts a Turing-passing voice agent could arrive as soon as this year

Takeaway: ElevenLabs’ single-minded focus on audio has allowed it to lead in expressive, high-quality voice synthesis and positioned it to define how we’ll talk to technology next.

Intron, a Nigerian AI startup, is expanding its voice technology from healthcare into finance, telecom, and legal sectors.

The company has introduced three core models:
• Sahara-Optimus – speech recognition for African accents
• Sahara-TTS – TTS with 80+ voices and 40+ accents
• Sahara Voice-Lock – voice authentication to fight fraud

Use cases are already emerging. Nigeria's Ogun State Judiciary uses Sahara for courtroom transcription, reducing session time. Rwanda’s Ministry of Health is rolling it out for medical record documentation.

Key points:
• Expansion into finance, telecom, and law
• Three new voice AI models
• Adoption by judiciary and health ministries
• Focused on African voices and dialects

Takeaway: Intron is building critical voice infrastructure tailored for African users, aiming to serve startups, enterprises, and governments across the continent.

ChatDeutsch has launched a free, no-login conversational AI platform for German speakers. Powered by OpenAI’s GPT-4.1 Nano, the site enables native-language chat with no barrier to entry.

It’s designed for simplicity: users visit the site, ask a question in German, and receive an instant response. Use cases include drafting emails, translating concepts, and answering general knowledge questions.

Key points:
• Free AI chat in German
• No registration required
• Uses GPT-4.1 Nano
• Accessible and localized

Takeaway: ChatDeutsch.de lowers the barrier to AI access for German-speaking users, offering a fast, easy-to-use tool for everyday communication.

🎙️ Mic Drop

What else is making noise in voice AI.

• ElevenLabs expands global presence to grow support footprint → pymnts.com
• ScienceSoft debuts HIPAA-compliant voice scheduling assistant → aithority.com
• Ultatel launches no-code enterprise voice AI agent → streetinsider.com
• Canary Technologies adds voice AI tools for hotel guest experience → hospitalitynet.org
• LiveOne and Synervoz partner on B2B voice AI growth → finance.yahoo.com
• Roundup of best AI voice generators for agencies → 9meters.com
• SpeechSSM enables hour-long AI voice narration → opentools.ai
• Build conversational search with Amazon OpenSearch → aws.amazon.com
• Android TTS guide with playback controls → medium.com
• How to run Dia-1.6B, an open-source ElevenLabs alternative → vocal.media
• Silverchain pilots voice AI in home aged care → healthcareitnews.com
• RACGP issues conversational AI guidance for doctors → racgp.org.au
• Tips for integrating AI voice in business workflows → techrepublic.com
• SpeechSSM explores audio-native LLMs → techxplore.com
• KAIST unveils SpeechSSM for always-on speech-native AI → miragenews.com
• CMSWire on adoption hurdles for voice AI in contact centers → cmswire.com