Contxto – There’s no doubt that English is a dominant world language. However, the ratio of quality English content versus Spanish leaves much to be desired. In turn, this denies non-English speakers certain resources and important information. Anyone who doesn’t know English is essentially missing out.
To tackle this issue, the Mexican VC fund ALLVP partnered with Slang, an edtech startup teaching technical English. Together, they want to allure Latin American entrepreneurs into the startup and VC realms by helping them improve their English skills. Not only that but ALLVP also invested US$2.5 million into the company.
Last Tuesday, I attended a Mexico City press conference where ALLVP and Slang announced this partnership to strengthen Latin American founders’ English abilities. Slang, a U.S.- Colombian startup, creates English courses. Unlike others, though, it defines jargon and technical terms specific to different careers.
Slang creates niche courses ranging from English for healthcare, to law, finance and technology. Recently, they also launched a new free course covering the most important terminology in the startup ecosystem. Now users can learn the definitions of lean methodology, product-market fit, or even term sheets. Moreover, they can even start implementing these items in their day-to-day.
Slang declined to comment on the round’s details, waiting to release a proper statement in the upcoming months.
“Humanity managed to practically eradicate illiteracy,” said Diego Villegas, Slang co-founder and CEO. “There are very few people that don’t know how to read or write. Nevertheless, there’s a new type of analphabetism, and this comes with the internet era. I call this modern analphabetism or analphabetism 2.0.”
Before launching Slang, Villegas was a Colombian engineer earning his MBA at the Massachusetts Institute of Technology. With English as his second language, though, many of the case studies were confusing. It wasn’t until 1993 when he created MASA, an energy project developer using specialized English, that he came up with a product proposal.
“I realized that there was not a solution focused on the professional and business needs to learn English,” said Villegas. He founded Slang in 2013 with machine learning expert Kamran Khan.
Latin American founders not only deal with startup issues but also legal, political, social, economic regulations, plus a highly anglicized system. While English thrives on the internet, and the most valuable content available, Latin American founders face frustrating barriers. Even though they can access this information, many cannot understand the actual content.
I held an exclusive talk with Slang founders and found out some amazing things about the company. If you thought Slang was just an edtech company, you got only one side of the story. Slang is actually an edtech and a data company.
On one hand, the startup enriches the community by facilitating specialized learning support to people across various fields. Yet consider the potential cases of the generated data, translations and byproducts of the consumer-facing business unit.
“We’ve never told any other outlet about our strategic plans in the future, but here’s a little preview of our long-term plans,” said Diego Villegas when I asked him about the potential use cases for the data they’re going to generate.
According to the founders, they are potentially open to licensing their lexicon or vocabulary database in the mid to long-term. Bigtech could especially utilize the product, specifically its speech recognition technology and Natural Learning Processing (NLP) systems.
This is especially interesting for companies such as Google and Amazon, which are extremely efficient when dealing with conversational English. Nonetheless, when it comes to reading a research paper or technical literature, their algorithms would most likely err. Big time.
Not only that, but the lexicon would also be able to create a network of specialized English vocabulary, creating deeper and more implicit relationships between terms and words.
Back to the user-facing front, their ML algo would also help users customize their learning process. This means that the platform would be able to identify the most difficult words to pronounce for certain users. Also, it will detect the ones that they keep forgetting the most in order to adapt and exponentially reduce the learning curve.
Origins: The startup began in Boston, Massachusetts as an experiment to blend Machine Learning and Natural Language Processing into one useful product. In the early years, it participated in MassChallenge where it raised US$30,000 of accelerator funds.
Funding: To date, Slang has reportedly raised a total of US$2 million from Founders Capital, one of Silicon Valley’s largest VC firms. We are still waiting for confirmation over this amount, however. Apart from that pre-seed round, it recently closed the deal with ALLVP, yet to be known.
History: Slang first started operations in Colombia where it developed, tested and iterated the original versions of the product. Following this, it expanded to Mexico and Chile. Today, it currently offers over 60 courses with 30 employees.
Goals: Its ambition, however, is to develop 1,000 courses by the end of the year with 60 employees, the majority based in Mexico. There are also talks to introduce Slang to Latin American governments so that employees develop working-proficiency English. All of this comes after years of hard work as well as trial and error.
“Our first course took around a year to complete,” said Cameron, Slang’s CTO. “Now, it takes around four weeks to fully develop a new course, and all of this has been possible thanks to our Machine Learning processes and our feedback loops with customers. What’s more, it is really important to understand the potential of the data in terms of word correlations.”
Apart from the money, ALLVP is also contributing to the platform by providing insights, guidance and quality control concerning this new course targeted to founders.
“The algorithm doesn’t change, the course creation process is the same across all subjects,” clarified Villegas. “Nevertheless, we always need to delimit the range of subjects to be covered, and the reassurance from experts to make sure this is relevant content.”
Nowadays, English-language students can learn between 10,000 to 20,000 words over Slang’s platform. With the traditional system, users learned an average of 2,000 words over the same amount of time. Such improvements are due to the startup’s adaptive database, meaning that users don’t waste time reviewing words they already know.
Stay tuned for the updates regarding the funding round.