ALL-IN Podcast Live: Mati Staniszewski, Co-Founder & CEO of ElevenLabs | RAISE Summit 2026

ElevenLabs Financial and Operational Growth

• Company revenue reached $600 million ARR, up from $350 million in 3–4 years and $100 million after the first 20 months of post-launch growth.
• Revenue trajectory accelerated: 10 months to reach $200M ARR, followed by 5 months to $300M, closing the previous fiscal year at $300M before jumping to $600M.
• Workforce expanded to 600 employees while maintaining a 100% retention rate for the original 10-person core research and engineering team.
• Revenue growth reflects the expansion from a research-heavy startup to a full-stack communication platform integrating audio research with go-to-market functions.

Organizational Structure and AI-Driven Management

• The company eliminated traditional Product Manager roles, optimizing for generalists who can code, understand design, and know the customer, supported by AI tools.
• Engineering talent is embedded into non-engineering departments (HR, Legal, Sales) to automate processes and enforce AI security checks on user-generated content.
• Internal teams are structured as small, autonomous units of 5–10 people optimized for specific industries like telecom, financial services, and healthcare.
• The company utilizes its own AI voice agents for inbound sales (inbound AI SDRs), enabling faster data capture and higher customer information quality compared to web forms.

Consumer Behavior and Interaction Shifts

• Enterprises now prefer AI operators over human call centers due to the ability to interrupt conversations naturally, high fidelity, and lack of small-talk friction.
• Consumers exhibit a behavioral shift from reactive inquiry to proactive assistance, where agents anticipate needs based on past interactions and real-time context.
• Users report greater emotional openness with AI agents regarding sensitive topics (e.g., debt collection), overcoming the shame barriers often present in human interactions.
• Input methods are evolving; users increasingly employ hardware pedals or voice-streaming techniques to feed "stream of consciousness" prompts to LLMs, bypassing typing fatigue.

IP Protection, Legal Safeguards, and Celebrity Partnerships

• ElevenLabs employs three core safeguards: full traceability of generated content, moderation on both voice and text levels for commercial/scam attempts, and AI detection tools for upload samples.
• The company launched a voice marketplace where talent can authenticate and license their own voices, having returned over $22 million to the community of voice creators.
• Paid partnerships with high-profile talent (e.g., Matthew McConaughey, Jamie Foxx) involve interactive content integration rather than static recording, allowing for dynamic, language-translatable interactions.
• A specific deal with the estate of James Earl Jones licensed the Darth Vader voice for Disney, enabling live, interactive interactions in Fortnite rather than static movie dubbing.

Strategic Positioning Against Frontier Model Giants

• The platform maintains model agnosticism, allowing customers to utilize Anthropic, OpenAI, Google, or open-source models without dependency on a single provider.
• Competitive advantage relies on specialized architecture, specific unlabeled audio data, and deep industry-specific workflow knowledge rather than raw model scale.
• ElevenLabs plans to invest more in proprietary research for interaction-specific models to avoid the risk of data leakage or being outcompeted by frontier model developers.
• The company does not view partnerships with larger AI labs as existential threats, citing the value of their specific voice and orchestration layer in the broader AI ecosystem.

Mission-Driven Use Cases and Future Outlook

• The company provides critical assistive technology for individuals who lost their voices to ALS or throat cancer, including a historic speech by Congresswoman Jennifer Wexton.
• Emotional restoration projects include helping a bride who lost her voice before her wedding to speak her vows again, demonstrating the technology's humanizing potential.
• CEO Mati Golan asserts that current voice AI has passed historical tests for Artificial Intelligence, though new metrics are needed to define true AGI.
• The forward-looking goal is to achieve a "voice Turing test" by the end of the year where interactions are indistinguishable from human conversation across all applications.