Sign language as infrastructure for half a billion

The problem nobody solved for

Muzamil opens the conversation by asking Ali about growing up blind in Pakistan. Ali’s response frames the core insight: “I spent my 32 years of my life without seeing anything. But pain comes when you’ve lost something. If I didn’t have that thing from the beginning, then it wasn’t as much pain.” This is the philosophical bedrock of what would become DeafTalk.

Ali holds an MBA and MPhil in supply chain management, worked on Telenor’s open mind training program, and was by any measure successful. But he noticed something the market had missed: while there were apps for blind users, nothing existed for the deaf community. Half a billion deaf people globally lacked access to certified sign language interpreters. In Pakistan alone, the number sits at 10 million.

“If I’m French, I’ll have a French interpreter,” Muzamil observes. “But if I’m deaf, how do I communicate with you?” The answer, until DeafTalk, was: you couldn’t. Not easily. Not at scale. This is infrastructure poverty.

How a hackathon became a startup

The founding story is not a Silicon Valley origin myth. It’s an accident. In 2017, Telenor organized a digital revolution initiative and asked employees to develop apps for persons with disabilities. Three co-founders submitted their idea: Ali and Vamek, plus Abdul Qadir, who is also blind. All three were working at Telenor. They won the competition, received cash grants, and in the euphoria of victory, Vamek had a different realization.

He was crying on stage. His co-founders didn’t understand why until he told them: “You both are my best friends because when he was in fifth grade, he was kicked out of school and he went to the US to complete his education in engineering. And he became the first deaf person who did software engineering and came back to Pakistan only to realize that the problem is still there.”

Vamek, born deaf, had lived in Pakistan through fifth grade with no speech therapy infrastructure. His family sent him to the US where he learned to speak, to code, and to measure the gap that separated Pakistan from the deaf-accessible world. He came home to find nothing had changed. And so on April 9, 2018, exactly the day they were incubated at NIC (the National Incubator Center), the three co-founders quit Telenor to build DeafTalk.

“The first 3-4 days, it was really shocking,” Ali recalls. “We were married, and now we have everything left and left. But it gave us the conviction that we have to open everything up.”

The product: 30 seconds, not weeks

DeafTalk operates as a marketplace. Deaf users download the app, connect to a certified sign language interpreter via video call, and communicate their needs in real time. A doctor’s visit, a bank transaction, an Uber driver exchange. The interpreter translates sign language to spoken language and back. The entire call takes minutes. The entire wait time takes 30 seconds maximum.

This is revolutionary only because the status quo is broken. In the US market, booking an interpreter requires 4-5 days notice. A 15-minute call costs 45,000 rupees minimum because interpreters charge for a 3-hour minimum booking. An emergency? Impossible within that same week.

DeafTalk charges 100 rupees per session. An interpreter working 4 hours a day can make 35,000 rupees monthly. One certified interpreter handles 40 calls a day, which means the unit economics stay positive even as the company scales. The app is 24-7. No minimum booking. No advance notice.

The numbers prove the market exists. By the time of this episode in April 2021, DeafTalk had 19,000 monthly active users across Pakistan and Singapore, with 6,000 daily active users. The founders had already generated 3.5 million rupees in revenue. They were onboarding 300+ new users daily through pure word-of-mouth.

Community closes the loop

Muzamil asks the obvious question: how do you reach a population with no database, no registry, no census of who is deaf? The answer reveals something about Pakistani society that data-driven Silicon Valley often misses.

“This community is very closely aligned with each other,” Ali explains. “They could only understand each other’s. Friends are very strong.” The social fabric of Pakistan’s deaf community is tight enough that one user brings ten friends. No paid acquisition. No viral loops. Pure referral.

This is how DeafTalk grows. Muzamil asks Vamek directly whether AI will eventually replace human interpreters. The answer is yes, but not yet. The technical challenge is that sign language is three-dimensional. Most smartphone cameras record in one or two dimensions. But Vamek and his team are already working with Google and Facebook on machine learning models that can translate sign language to voice in real time.

“We would love to be the Google Translator for the deaf community across the globe,” Vamek says. In one or two years, he expects the beta version to launch. Until then, human interpreters remain the bridge.

Barriers that software cannot fix alone

Accessibility requires more than an app. Muzamil points out that he’s watched DeafTalk on YouTube with subtitles available, yet most of the deaf community cannot read those subtitles. Why? Because 99% of deaf people globally cannot read or write the native language. Their education is conducted in sign language. Their literacy is in sign language.

This is why Pakistani universities like Virtual University have already integrated DeafTalk into their course offerings. If a course is taught in sign language interpretation, then deaf students can become software engineers, accountants, and financial experts. If not, the course remains inaccessible.

The structural barrier is policy. In Pakistan, there is no legislation that protects the rights of deaf citizens the way US and UK law does. But something is moving. Asad Umar, the speaker of Pakistan’s National Assembly, has a daughter who is deaf. He has mandated that the entire National Assembly website be available in sign language. This is policy following product, not the other way around.

Where Pakistan will lead

When Muzamil asks where Ali sees Pakistan in ten years, the answer is unhedged: “I am 100% sure that not today, Pakistan will be the leading country in technology, will be the leading company in the businesses. Five or six years ago, no one talked about businesses. Now every child who comes to university wants to run a startup.”

Vamek echoes the optimism: “Within ten years, I believe Pakistan will be one of the most, and I’m trying to hard to market because we have a lot of data. With everyone working together for the empowerment and the betterment of Pakistan, I’m really hopeful for that.”

Muzamil wraps by encouraging viewers to download the app if they know someone deaf. The link is in the description. Go to Google Play or Apple App Store, log in in three steps, start using it, send feedback. That’s how infrastructure gets built in markets nobody else wants to serve yet.