How Myanmar's Coup Fueled an AI Bet for Migrant Women

In February 2021, Myanmar's military ended a decade of tentative democratic opening with a coup, and within two years drove the country's tourism industry into a deeper collapse than even 60 years of civil war had caused. For Byron Hartzler, an entrepreneur who spent nearly a decade building trekking, rafting, and homestay routes through one of Myanmar's most remote regions, that collapse meant selling the company. For Matt Wallace, founder and CEO of the Myanmar-based nonprofit ONOW, the same period of crisis proved something he had already started to suspect: the fastest way to help more entrepreneurs isn't hiring more coaches, it's building software that frees the coaches he already has from data entry.

Both men spoke with Students Incorporated about what happened next, and their two stories, one of a business that couldn't survive a coup, one of a nonprofit that used automation to grow through one, turned out to be the same story from opposite sides.

An "Off the Beaten Track" Bet That Paid Off, Then Didn't

Hartzler, who is Canadian and American and has lived in Asia since 2011, started Myanmar Adventure Outfitters in 2015, as the country reopened to foreign investment and tourism after decades of isolation. He chose a region with no existing tourism infrastructure at all: villages of 30 to 40 houses where he was, by his own account, the first foreign person many residents had ever met. "We chose to live in an off the beaten track part of an off the beaten track country," he said. Building the business meant explaining the entire concept of tourism to farmers who had never imagined a stranger would pay to photograph their rice fields.

It worked. By the company's peak, Hartzler's guides ran mountain biking, rock climbing, river trips, and village stays, and profits were doubling year over year as Myanmar's tourism numbers climbed alongside the country's broader opening. That growth also pulled Hartzler and his team into development work: education sponsorships to keep teenagers out of exploitative factory jobs, water filtration, and health education.

Then, in his words, "COVID was the great tourism killer of the world." The numbers bear that out. Myanmar had welcomed more than 4.3 million foreign tourists in 2019, according to industry data compiled by Standard Insights; foreign visitor spending that year reached $2.84 billion, and tourism accounted for roughly 5.9% of GDP. By 2020, arrivals had fallen to about 903,000 and visitor spending to $460 million. Then came the coup. According to reporting from Frontier Myanmar, international arrivals for 2021 totaled just 130,947, a fraction of a fraction of the pre-pandemic peak. The International Labour Organization estimated the country lost 1.6 million jobs that year between COVID and the coup combined, and the tourism and hospitality sector alone shed an estimated 27% to 31% of its jobs after the military took power, per ILO figures reported by Al Jazeera.

Hartzler's account of the mechanics matches that collapse exactly. The area his company operated in has a "not so distant, fairly recent history of civil war," he said, and after the coup, "there's just no tourism, like almost zero in the country now." By the time of the interview, he and his partners were selling the company and moving to Thailand.

Diaspora as the Next Business Plan

That move wasn't just an exit. It was a redirection toward a market the coup itself had helped create. Years of conflict and economic collapse have pushed a large and growing share of Myanmar's population into Thailand. The International Organization for Migration estimated in September 2025 that Thailand hosts roughly 4.6 million Myanmar nationals in regular and irregular status combined, and IOM has reported that nearly 700,000 more entered the country in the first four months of 2026. Registered migrant workers alone rose from about 3.06 million in 2024 to 3.65 million by October 2025.

Hartzler has deep relational ties to one part of that diaspora, the Shan people, built over a decade of running tours through Shan State. Wallace's ONOW had already been eyeing an expansion into northern Thailand to serve that same community. "It makes a ton of sense for Byron to come along with us," Wallace said, "and even show us a little bit different way to do it with that community in particular." What began as two friends who'd known each other since their kids attended school together in Yangon became a merger of Hartzler's community trust and ONOW's delivery model.

The Chatbot That Ate the Admin Work

That delivery model is where ONOW's story gets surprising. ONOW exists to help young migrant and refugee women, largely factory workers, become "financially capable and empowered entrepreneurs," in Wallace's words: managing money safely and eventually launching their own small businesses. Before the pandemic, that meant intensive one-on-one coaching for about 300 women. Wallace said the team quickly learned something counterintuitive: "This is not a training. It can't be a training. You can't really train business. It needs to be experienced."

But experience at scale required automating everything that wasn't the actual coaching relationship. ONOW built its own chatbots inside Facebook Messenger, reasoning that the women it served, many working 12- to 16-hour factory shifts six days a week, needed to learn financial and digital literacy on their own phones, on their own time. Those chatbots have since been used by more than 300,000 people, according to Wallace.

The bigger shift came on the operations side. ONOW discovered its coaches were spending 28 of their 40 weekly hours on reporting and call prep rather than actual coaching. By routing that data collection through automated tools instead, coaches could spend a 45-minute call on strategy instead of intake. The result, per Wallace: ONOW went from supporting roughly 300 businesses before the pandemic to more than 10,000 with real-time, individualized coaching.

That trajectory checks out against independent reporting. Youth Business International, which helped fund ONOW's early digital pivot through a Google.org-backed COVID recovery program, reported in a November 2020 interview that the chatbot shift had already taken the organization from 300 businesses to roughly 13,000 reached through the chatbot alone. By October 2023, Forbes contributor Anne Field reported ONOW had coached more than 50,000 businesses, with many entrepreneurs seeing earnings more than double.

"There's very little that is as demonstrative of agency than a business owner choosing to launch her own business."

That scaling problem, reaching women who are hard to reach through conventional apps, is a documented industry-wide gap, not a Myanmar-specific one. GSMA's 2025 Mobile Gender Gap Report found the gender gap in mobile money ownership has widened since 2021: women in low- and middle-income countries are now 36% less likely than men to own a mobile money account, up from 30% in 2021. GSMA estimates 440 million more women across South and East Asia and the Pacific need to adopt mobile internet by 2030 to close the regional gap. ONOW's workaround was to meet women where they already were: inside an app, Facebook Messenger, that Wallace said nearly everyone in Myanmar with a smartphone already used, rather than asking them to adopt something new.

What They're Betting the Rest On

Both men described the same underlying discipline for surviving repeated crises: know the mission, and be willing to kill anything, a product, a service, even a country's operations, that stops serving it. Wallace pointed to what he called the "sunk cost fallacy," the temptation to keep funding a program because of what's already been invested in it. "If that product or service isn't achieving the impact you need, you have to kill it," he said. "You have to kill it fast." Hartzler, for his part, credited plain persistence: "None of what Matt and I have done in Myanmar happens unless you have some drive to just keep on pushing when everyone else says no."

ONOW's next bet is to stop being the only organization running this model. Wallace said the technology his team built is now being given to other organizations so they can replicate ONOW's coaching-plus-automation approach with their own populations, while ONOW itself expands geographically, starting with the diaspora community Hartzler now understands better than almost anyone else in the room.

← Back to all posts