Speech recognition startup pulls in a respectable series B round of funding
There still is significant funding heat to be found in biometrics. A speech recognition startup is reporting that it has closed a $72 million series B round of funding after pulling in $47 million in equity funding this week.
According to Crunchbase, Deepgram has raised $86 million in nine rounds. The company had a post-money valuation of $100 million to $500 million in February 2021, according to business data aggregator PrivCo.
Deepgram’s CEO, Scott Stephenson, says the company is building AI models for understanding speech. The company is taking an audio-first approach to its code rather than the textual approach found in natural language processing.
Company executives claim their algorithms have transcribed 1 trillion words derived from 10,000 years of raw audio.
Early-stage venture fund Madrona Venture Group is considered the leader of the $47 million raised. The firm’s portfolio has a solid AI segment, but is light with voice recognition. In 2018, it invested in Pulse Labs, which makes products companies use to create voice apps.
Hedge fund Alkeon Capital Management also participated in the last installment.
Citi Ventures invested in Deepgram in two pieces for the series B. The fund is a corporate venture unit within financial services giant Citigroup, which has had its own voice recognition ambitions over the years.
The global market for voice algorithms like Deepgram’s is experiencing moderate growth. A recent analysis by Grand View Research predicts a compounded annual growth rate of 15.3 percent between now and 2030.
Grand View estimated the market size was $14.4 billion in 2021.
Meanwhile in another funding announcement for an AI development startup with some implications for biometrics providers, V7 has raised $33 million in series A funding, Business Insider reports.
London-based V7 helps automate segmentation and annotation for computer vision models, such as for facial recognition.