AssemblyAI, a leading provider of customizable artificial intelligence (AI) speech models, has raised $50 million in a funding round led by Accel. The new capital will help AssemblyAI accelerate research and development of advanced natural language processing models, while expanding go-to-market initiatives to meet rapidly growing demand.
With its innovative approach of developing highly accurate AI speech models and making them easily accessible via API, AssemblyAI enables companies to quickly integrate next-generation speech technology into their products and workflows. Over 4,000 brands now leverage AssemblyAI to process vast amounts of audio data - currently handling around 25 million API calls and 10 terabytes of data daily.
AssemblyAI co-founder and CEO Dylan Fox started the company in 2017, seeking to democratize access to superior voice AI that major cloud providers were not offering. He envisioned a “Stripe for AI models” - where developers could seamlessly access cutting-edge speech recognition through a simple API.
The flexibility of AssemblyAI’s platform allows customers to build customized speech products for meeting transcription, content moderation, speaker identification and more. For example, mobile meeting assistant app Fireflies harnesses AssemblyAI to transcribe calls, podcasts and video meetings.
Fox asserts that AssemblyAI’s proprietary models significantly outperform competitors in accuracy, capabilities and new feature integration. He credits the 200% year-over-year customer growth to the platform’s speed of innovation.
Now armed with $115 million in total funding, AssemblyAI aims to accelerate research into what Fox calls “a universal speech model” trained on over a petabyte of voice data. The company also plans to grow headcount by 50-75% in 2023 to keep pace with intense product demand.
According to Fox, “We have years of runway thanks to the new funding round, and are seeing an incredible amount of demand and product adoption given the mainstream push around AI.”