Compose.ai raises US$2.1 million to build its own language model that would help others write faster. Compose.ai intrinsically is an auto-complete that works while web browsing.
The company is also developing an artificial intelligence, which would recognize and learn the voice of the user in order to provide a quicker and better response. Landan Sanford, Co-founder of Compose.ai, said that he believes in the coming five years, people won’t type every word they want to write.
Compose.ai is developing a Google Chrome extension that the customers can access without the need to get corporate permission. While explaining the working for their algorithm, Sanford described it as a multi-tier product.
Read More: Amazon India Launches ‘ML Summer School’ For Students
The first tier learns from reading the web itself, learning English. The second tier deals with particular domains, like email. The third tier learns a user’s voice, and the fourth tier will be related to a company’s generalized approach to language.
It is a very interesting concept as this would enable the company to provide its workers a common language model that could offer word-choice and linguistic preference when they write. After the funding, the company is planning to hire a few experienced engineers to work with them.
Compose.ai Co-Founder Shuffett said that they wish to stay small for now. The start-up currently has three full-time employees and a few contractual workers. The product Compose.ai wants to create is not something simple. The founders said, “the personalization work that its product will execute for customers can be expensive if done incorrectly.”
They also added that they could, in the future, introduce an economical US$10 monthly subscription, but if executed in a naive fashion, the company might have to spend US$500 in technical costs to support the same subscription. But Compose.ai is confident that their subscription plans will perform well in financial terms.