Slang.

A bank teller in Bogotá is learning to say "non-performing loan." Correctly.

She is not learning to order coffee in London or describe her weekend in the past tense. She is learning the exact English her job demands - the phrases a credit committee uses, the way a risk report is worded. The app on her phone is Slang. And it knows she works in banking, because that is the entire point.

Slang is a Boston-headquartered edtech company that teaches professional, industry-specific English to the global workforce. Not English in general. The English of engineers, doctors, logistics managers, and yes, bank tellers. It runs what it calls the largest catalog of specialized professional English courses in the world: more than 200 of them, sorted by industry and role, delivered through a mobile-first platform that companies and universities buy on behalf of the people who need it.

Most English apps are very good at teaching you the wrong thing.

There is no shortage of ways to learn English. Apps with cartoon owls, courses with cheerful dialogues, classes that drill you on the present perfect. They are pleasant. They are also, for a working professional, slightly beside the point. A petroleum engineer does not need to discuss her favorite color. She needs to read a drilling report.

This is the gap Slang exists to close. Conversational English gets you through a dinner. Professional English gets you through a career - a promotion, a client call, a research paper, a contract. For millions of skilled workers, particularly across Latin America, the barrier between their ability and their opportunity is not skill. It is vocabulary they were never taught, because the textbooks stopped at "How are you today?"

The irony is hard to miss. The people most in need of advanced, technical English are often the ones the language industry serves last - because their needs are specific, fragmented, and expensive to build for. One course for nurses, another for auditors, another for civil engineers. It is far easier to sell everyone the same lesson about hobbies. Slang decided the hard, fragmented version was the actual job.

It started because the founder couldn't read his own homework.

In 2013, Diego Villegas was pursuing an MBA at MIT and kept hitting the same wall: the specific, technical English of his business cases and course materials was a thicket. He spoke English. He still couldn't fully follow it. If that was true for him - an MBA candidate at MIT - what about everyone without that runway?

On campus he met Kamran Khan, a computer science student finishing a master's with a focus on artificial intelligence, who had just won an iOS app development competition. Villegas became CEO; Khan became CTO. Their bet was that AI and natural language processing could make learning a language radically more efficient - and that the most valuable place to aim that efficiency was not at tourists, but at professionals.

What began as a research project became a company, and the company kept the research habit. Slang's pitch is not "learn English." It is "learn the English you, specifically, will be paid to use." That precision is the product.

Pick your industry. The English follows.

The Slang platform looks, at first, like any well-built learning app: short lessons, listening exercises, a progress ring, the satisfying ding of points earned. Look closer and the difference shows up in the menu. Instead of "Unit 4," you choose Environmental Engineering, or Financial Services, or Healthcare. The lesson then teaches you to say things like "our firm is committed to sustainable construction" - because that is a sentence you might actually have to write.

Underneath the friendly interface is the part the founders care about most: an adaptive engine built on machine learning and NLP that decides what you see next based on your role, your level, and what you keep getting wrong. Companies use it to push a sales team and a legal team down completely different roads. Universities use it to make sure an engineering graduate leaves knowing engineering English, not just English.

Investors with a thesis, customers with a payroll.

Belief is cheap; cheques are not. In 2019, ALLVP put in around $2.5M. In December 2021, Slang closed a $14M Series A co-led by DILA Capital and ALIVE Ventures (Acumen Latam Impact Ventures), with Salesforce Ventures, Roble Ventures, and Impact Engine joining. That is a notable roster for a company teaching the subjunctive to auditors - it signals that serious money sees professional English as workforce infrastructure, not a nice-to-have.

The customer side is just as telling. Slang's corporate clients span financial services and banking, pharmaceuticals and healthcare, consulting, technology, consumer goods and retail, logistics, mining, and oil and gas - the kinds of industries where one mistranslated clause has consequences. Universities use it to make graduates job-ready. The common thread: these buyers do not want their people to chat. They want them to work, in English, in their field.

Language should never be the thing standing between talent and opportunity.

Strip away the courses and the funding and Slang's mission is stubbornly simple: close the professional English gap for the global workforce. Not because English is superior, but because it is, for now, the operating language of global business - and a skilled professional should not lose a promotion to a vocabulary list.

That mission explains the choices. Why specialize by industry? Because that is where the gap actually bites. Why sell to companies and universities? Because that is where access can scale fastest. Why anchor the team across Colombia, Mexico, and Brazil while keeping a Boston address? Because the people Slang serves most are there, and a company that forgets its users tends to forget its point.

The AI that translates everything will not make this go away.

Here is the obvious objection: machine translation is getting frighteningly good. Won't software just speak for everyone soon? Maybe, in the margins. But the bank teller who can speak the language herself, in the room, in the meeting, will always move faster than the one waiting for the subtitles. Fluency is not a fallback. It is leverage.

If anything, the same AI wave that threatens generic translation strengthens Slang's hand - the better its NLP gets at understanding how an industry actually talks, the sharper and more personal each learning path becomes. The company that started because one MIT student couldn't read his own case studies is, increasingly, a quiet piece of how the rest of the workforce reads theirs.