One of the most challenging aspects of language acquisition is developing speaking skills, particularly without consistent practice opportunities. ELSA is an application designed to address this difficulty by utilizing speech recognition technology to refine pronunciation. Headquartered in both San Francisco and Ho Chi Minh City, ELSA has recently announced the completion of a $15 million Series B funding round, spearheaded by VI (Vietnam Investments) Group and SIG. Additional investors included previous backers such as Google’s Gradient Ventures, SOSV and Monk’s Hill Ventures, as well as Endeavor Catalyst, Globant Ventures and U2’s The Edge (David Evans).

These newly acquired funds will be allocated to broadening ELSA’s presence in Latin America and constructing a robust B2B platform. This platform will enable businesses and educational institutions to extend the app’s coaching capabilities to their workforce or student body. Established in 2015, ELSA, which represents English Language Speech Assistant, currently boasts a user base exceeding 13 million. The company’s prior funding round was a $7 million Series A, secured in 2019.

Beyond Latin America, ELSA will also concentrate on growth within Vietnam, India, and Japan, all of which demonstrated significant demand in the previous year. The company has recently established a collaborative partnership with IDP, a key stakeholder in the widely recognized IELTS English language test, which now suggests ELSA as a resource for test preparation. ELSA is also collaborating with Vietnamese language academies such as IMAP and Speak Up, the online learning platform YOLA, and corporate entities including Kimberly Clark, Intel and ATAD.

Vu Van, ELSA’s co-founder and chief executive officer, shared with TechCrunch that a primary driver for users is the desire to enhance their English speaking abilities to unlock career prospects and increase their potential earnings. In countries like Vietnam, India, and Brazil, individuals with strong English communication skills can earn approximately two to three times more than their peers, she explained.

“This incentive fuels considerable demand within our English learning community in Vietnam, India, and Brazil, particularly during the COVID-19 pandemic, when we observed a substantial surge of interest from the Latin American region,” Van stated.

In Vietnam, where Van originates, English learners frequently dedicate a significant portion of their discretionary income to online or in-person English instruction. “However, a large number of English learners continue to struggle with improving their speaking ability because others have difficulty understanding them or they experience apprehension when attempting to speak,” she noted. ELSA was created to provide them with an easily accessible tool to improve their pronunciation and build confidence in spoken English.

Other applications focused on English pronunciation include FluentU and Say It. Van highlighted that a key differentiator for ELSA is its exclusive voice recognition AI technology.

“The distinctiveness of our AI lies in the extensive collection of accented English voice data gathered from millions of users, which we have leveraged over the past several years to train our AI model. This results in greater accuracy in recognizing and comprehending non-native English speakers globally,” she explained. “In contrast, other available voice recognition technologies may perform well with native speakers but encounter challenges in understanding non-native accented English learner communities.”

Rather than offering feedback on isolated words, ELSA’s application also refines individual sounds and provides users with comprehensive insights on how to refine their pronunciation, encompassing “sophisticated prosodic speaking elements like intonation, rhythm and fluency to help them speak English more naturally, a feature not commonly found in competing products,” Van added.