Mati Staniszewski on ElevenLabs: Voice AI for Human-Technology Interaction

ElevenLabs' Mission and Growth

• 🚀 ElevenLabs was founded to transform how humans interact with computers and each other through voice.
• 📈 The company has achieved rapid growth, reaching 350 employees globally and an annual recurring revenue (ARR) of $300 million, split evenly between self-serve and enterprise clients.
• 💡 They build foundational audio models for creating human-sounding speech, understanding speech, and orchestrating interactive voice experiences.

From Dubbing to Foundational Models

• 🇵🇱 The initial inspiration came from the poor quality of Polish movie dubbing, where a single voice narrated all characters, highlighting the need for high-quality, emotional voice translation.
• 🧠 Recognizing limitations of existing models, ElevenLabs invested in deep research to develop their own foundational audio models, focusing on architectural breakthroughs rather than just scale.
• 🛠️ Their strategy involves creating specialized "labs" with researchers, engineers, and operators to tackle specific problems, sequencing research first before building product layers.

Elevating Voice Quality and Personalization

• 🗣️ Early models produced robotic speech, prompting a focus on making voices sound genuinely human and expressive.
• 🎯 ElevenLabs provides voice coaches to help enterprises select the ideal voice for their specific use case, language, and brand identity.
• ✨ The future envisions personalized and dynamic voice choices, adapting to user preferences, time of day, or even emotional state.

Transformative Agent Platform Use Cases

• 📞 The agent platform is shifting customer support from reactive to proactive assistance, as seen with Misho's e-commerce agent guiding users through shopping.
• 🎮 It enables immersive media experiences, such as interacting with Darth Vader's voice in Fortnite, bringing characters to life.
• 📚 Education is a key focus, offering personal AI tutors for subjects like chess with grandmasters or practicing negotiation with an FBI expert.
• 🏛️ ElevenLabs is also collaborating with Ukraine's Ministry of Transformation to create an "agentic government" for citizen services and education.

Future of AI and Human Interaction

• 🤖 While AI companions will be prevalent, the speaker is more enthusiastic about "super assistants" that proactively manage daily tasks and provide personalized information.
• 🎓 AI tutors are expected to revolutionize education, offering highly personalized learning experiences, though human-to-human interaction will remain crucial.
• 🗣️ Voice is seen as the ultimate interface for controlling technology, from smart devices to robots, allowing for more natural and immersive interactions in daily life.