Through the AI trends staff
Following the identity of the speech, progress in the AI is developing in the market, attracting venture capital and funding startups, and challenging the established players.
According to the recent account, according to a global complex research estimate by 2025, the market is to advance the market through the increasing acceptance and use of speech identification devices. Analytical insights. Better speed and accuracy include the benefits of developing technology.
San Francisco’s assembly, a new growth company, is offering an API to identify speech that is able to imitate videos, podcasts, phone calls, and remote meetings. The company was founded by CEO Delin Fox in 2017 and has been baking from the Wi -Combinite, a startup accelerator, as well as NVIDIA.
Fox has an extraordinary background for a high -tech entrepreneur. She is a graduate from George Washington University with a degree of business administration, business economics, and public policy. He got a job as a software engineer for machine learning in Sisco’s emerging product lab in San Francisco, worked on deep nerve networks and machine learning. It found the idea for the assembly and attracted capital from the Wi -Combinite, which helps to hire data scientists and data engineers so that they can remove the technology from the ground.
Asked in an interview with AI trends Fox said, “I taught myself how to program, which caused me to go on the way to learning the machine. He was working on Siri for Apple for Apple, Fox.
To accelerate work, Cisco wanted to get speech identification software. Fox was on the seat of Catebard for search. For example, “we saw the newborn,” was recognized as more market leader and more software owner than its competitors. (The acquisition of newborn in 19.6 billion will be finalized by the end of the year by Microsoft.) The young, emerging businessman was not affected. He said, “It was crazy that all the options were bad from the accuracy and the developer’s point of view.”
He was influenced by a San Francisco -based company, Tilio, founded in 2008, which issued a Tolio Voice API to make and receive phone calls hosted in the cloud this year. The company has then raised $ 103 million in venture capital. Fox said, “They were setting new standards for developers for good API.
Fox’s idea was to use AI and machine learning “to get the most accurate results, and make developers easy to add API to their products. A user call is a rail, calling call tracking and marketing analytical software, which includes NBC and Wall Street Journal to add NBC Journal to NBC and Wall Street Journal, NBC and Wall Street Journal. Intended.
“We are working to build as much human speech as close to the standard of identification,” said Fox. This has been a lot of work. ” He expects that this level will reach the plateau in 2022.
He targets companies that include speech identities in their products and make it easier to buy. Consumers pay on the basis of use. For every second of the audio copy, the assembly receives a portion of a penny. The client gets a monthly bill. If a user uses 10 hours a month, it costs about $ 9. If a user uses a million hours a month, the cost is about $ 900,000.
The sound recognition is a hot market. “Many new startups are being launched,” Fox said, providing the opportunity. “Many interesting new businesses are being built on voice data.”
Assembly products can detect sensitive topics such as hate speech and dishonor, so consumers can save human content moderation.
To explain what their technology difference is, Fox said, “We are an experienced team of deep learning researchers,” with the experience of companies, including BMW, Apple, and Facebook. “We create a very large, very accurate deep learning models, which have more accurate consequences than the traditional machine learning point of view. We make really big models using modern neural network technologies.” They compared the view that the Openi uses to develop his GPT-3 large language model.
In addition, they feature AI features above copy to provide audio and video content summary, which can be searched and configured. “It’s just beyond the copy,” said Fox.
The company currently has 25 employees and is expected to double in about four months. The business has been good. “There has been an explosion of online audio and video data, and users want to be able to take advantage of it, so we see a lot of demand,” said Fox.
Get more information Assembly.