ElevenLabs CEO: Why Voice is the Next AI Interface
Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
Podcast Summary
ElevenLabs CEO and co-founder Mati Staniszewski shares how his company ships research-grade AI at lightning speed, from text-to-speech and music generation to real-time voice agents. (02:18) The conversation covers their global hiring strategy, small autonomous team structure, and transition from creator-focused to enterprise platform. (24:45) Staniszewski explains how voice will become the next interface for human-computer interaction and details their Voice Marketplace success, which has paid creators over $10 million. (04:46) Key themes include maintaining startup velocity while scaling, navigating partnerships with creative industries, and building AI products that serve both individual creators and enterprise customers.
- Main themes: Scaling AI research at speed, global talent acquisition, creator-to-enterprise platform evolution, and ethical AI audio development through industry partnerships
Speakers
Mati Staniszewski
CEO and co-founder of ElevenLabs, Mati started the company between Warsaw and London with his co-founder and best friend of 15 years, Pyotr. He previously experienced the poor quality of voice dubbing in Polish cinema, which inspired the creation of ElevenLabs' advanced text-to-speech technology. Under his leadership, ElevenLabs has grown to 350 employees across multiple global hubs and raised significant funding from a16z.
Jennifer Li
Partner at Andreessen Horowitz (a16z), Jennifer Li has been working with ElevenLabs since their early days three years ago when the infrastructure team was just three people. She focuses on AI and enterprise software investments and has closely observed ElevenLabs' evolution from a small research team to a major AI audio platform.
Key Takeaways
Small Autonomous Teams Drive Speed
ElevenLabs operates with approximately 20 product teams of 5-10 people each, giving teams full independence to ship products quickly. (03:52) While this can lead to duplicative work or varying speeds, the high ownership model ensures rapid execution. Each team gets six months to prove their concept before it either continues or gets killed. This structure allows the company to maintain startup velocity even as they've scaled to 350 employees, proving that organizational design directly impacts shipping speed in AI companies.
Global Hiring Unlocks Hidden Talent
Rather than limiting recruitment to traditional tech hubs, ElevenLabs hires globally based on merit and passion. (08:05) They've discovered exceptional talent in unexpected places, including hiring a brilliant researcher who was working in a call center while developing open-source text-to-speech models. The company balances remote work with physical hubs in London, Warsaw, and San Francisco, allowing early-career employees to immerse in company culture while experienced remote workers maintain flexibility.
Partner With Industries Rather Than Disrupt Them
ElevenLabs' approach to working with creative industries focuses on collaboration rather than replacement. (13:32) They spent 18 months negotiating with major music labels like Merlin and Cobalt to create fully licensed AI music generation. The key was adding "forcing functions" with specific deadlines and demonstrating the technology directly to artists and label members to overcome initial resistance. This strategy of patient partnership-building has proven more effective than aggressive market disruption.
Voice Marketplace Creates Win-Win Economics
The Voice Marketplace allows creators to upload their voices and earn money when others use them, with nearly 10,000 voices now available and over $10 million paid back to the community. (14:34) This model solves the dual challenge of needing diverse voice options across languages and accents while providing creators with ongoing revenue streams. One Spanish voice that initially struggled in Spain became their top free voice in English-speaking markets, demonstrating the global reach potential of the platform.
Flatten Hierarchies to Maximize Impact
ElevenLabs removed job titles a year ago, creating a structure where tenure doesn't define hierarchy and new hires can immediately impact the company. (10:28) The flat organization consists only of division leads (research, creative, agents, sales, ops) with small teams beneath them. This approach ensures customers always get access to the best people regardless of seniority, and it allows smart, passionate individuals to elevate themselves quickly based on merit rather than politics.
Statistics & Facts
- ElevenLabs has paid over $10 million back to creators through their Voice Marketplace, which now hosts almost 10,000 voices across 70+ languages. (14:34) This demonstrates the significant economic opportunity created for voice creators in the AI era.
- The company has grown from a 3-person infrastructure team to 350 employees total, organized into approximately 20 product teams of 5-10 people each. (02:38) This scaling approach maintains startup velocity while building enterprise capabilities.
- ElevenLabs spent 18 months negotiating licensing agreements with major music labels including Merlin and Cobalt to create fully licensed AI music generation capabilities. (15:49) This timeline illustrates the complexity of building ethical AI partnerships with established creative industries.
