On the ground in Greece, MBA students test tech to help refugees

MarHub team
Intern Ramah Awad, Jerry Philip, EWMBA 19, and Sarrah Nomanbhoy, MBA 18, at the Ritsona Refugee camp on mainland Greece.

(Note: The first names of the refugees mentioned in this article have been changed to protect their privacy.)

In the sweltering June heat inside a refugee camp on the Greek mainland, Sarrah Nomanbhoy, MBA 18, listened closely as Khadija and her teenage son Tariq answered her questions about their family’s refugee odyssey.

Khadija’s family fled Syria’s civil war in 2015, heading for Turkey. A year later, they hired a smuggler who took them across channel to the Greek island of Chios in an overcrowded dinghy. They’d hoped to reach an uncle in Germany within three days by traveling through the Balkans, but found the borders had closed. Tariq searched his smartphone to figure out what to do next, but found that the information from humanitarian agencies too general, and that social media was filled with unverified rumors.

“They’ve now been waiting in Greece for over a year and a half,” Nomanbhoy said.

Nomanbhoy is not naïve enough to think that technology could remove the obstacles and traumas that this family and millions of other refugees on the move are facing. But she was there in the camp to find out whether she and her team—working on a startup called MarHub—could build a tool that could help refugees navigate the grueling asylum process.

“Lots of people have shown up in Greece with the ‘next app’ so there is a lot of skepticism,” says Nomanbhoy, a Silicon Valley native whose parents are both immigrants. “Refugees aren’t using these apps. We want to build something they will use.”

Over two weeks, three members of the MarHub team—which included Jerry Philip, EWMBA 19, and intern Ramah Awad—conducted interviews and tested their prototype with refugees in Athens, Ritsona, and in the Souda camp on the island of Chios.

Interviewing refugees

Souda camp
A view of the Souda refugee camp on the Greek island of Chios.

In interviews, the asylum seekers, many Syrian, but also Afghan, Pakistani, Nigerian and Moroccan, discussed the most challenging and uncertain aspects of the asylum process. Many shared that they were ill-prepared for the unexpectedly grueling asylum interviews. They were also largely unaware of their legal rights during the interviews, including the option to replace a translator or to review the interview transcript for inaccuracies. About 70% of asylum seekers receive negative decisions after this first set of interviews, and many are now in limbo pending the outcome of the appeal process, Nomanbhoy said.

Many face dire living conditions, violence within the camps, and racism from the local residents. “Witnessing the level of racism that asylum seekers are facing on a regular basis was really shocking,” says Nomanbhoy, recalling a few incidents where her team and asylum seekers they were interviewing were kicked out of local cafes on Chios.

Using iPads, the team showed refugees a prototype of the chatbot they developed, the first phase of MarHub’s crowdsourced information platform.

The chatbot allows a refugee to type questions through Facebook messenger and receive automated answers related to their asylum process based on where they are from, where they landed, and what their legal information needs are.

Creating a powerful tool

Marhub team
MarHub intern Ramah Awad (left), and Jerry Philip, EWMBA 19, (next to Awad) show MarHub’s prototype to NGO staff in the Ritsona refugee camp

It also enables them to connect to legal aid volunteers for additional assistance. As more refugees use the chatbot, the data collected will enable more accurate, timely responses, Nomanbhoy said. It is this crowdsourced information that will give the tool its power.

“Our goal is to help refugees feel informed and help them prepare for their asylum interviews,” said Philip, a product marketing manager at Cisco who grew up in Abu Dhabi and India and moved to the U.S. a decade ago. Philip says he and the rest of the team have all experienced the limbo of visa applications and understand the pain points and the tedious aspect of waiting.  “The hard part psychologically is not knowing how long the wait is going to be. Time moves very slowly when you’re waiting in a queue, but imagine not knowing whether you’ll be stuck for two months or two years.”

For Nomanbhoy, MarHub is a passion project that evolved from her interest in international migration. Growing up in the Bay Area, she saw immigrants driving entrepreneurship and making lasting contributions to the economy.

Before arriving at Haas, she launched a seed accelerator for startups in Sri Lanka and worked on a social enterprise to help migrant workers in Thailand send money home. Returning for an MBA was about getting the tools to start a company that could help change the way international migration is managed.

MarHub evolved from last year’s prestigious Hult Prize Challenge on Refugees. Nomanbhoy heard about the $1 million contest on a Friday night in November. By Monday she had met her co-founders—Philip; Peter Wasserman, MBA/MPH 18; and Srinivas Vaidyanathan, EWMBA 18.  They put together their pitch within a week. After edging out 50,000 entrants, the team made it to the last round of the competition’s regional finals. Encouragement from judges, support from their advisors, Assoc. Adj. Prof. Thomas Lee and UC Berkeley Law Prof. Katerina Linos, along with a $5,000 Dean’s Seed Fund grant and the $5,000 Hansoo Lee Fellowship, helped the foursome launch MarHub. Nomanbhoy also won a $12,500 Jack Larson Scholarship, enabling her to work full-time on MarHub over the summer.

The chatbot’s usefulness expands

After seeing MarHub’s prototype, many of the refugees asked the team how to sign up for the service. “The work feels really urgent now,” said Nomanbhoy.

The team expects to roll out a pilot service in partnership with Greece-based organization RefuComm in the fall and then go on to raise seed funding. The Haas students on the team have chosen courses this year—including Descriptive & Predictive Data Mining, Startup Sales, and Lean Product Management—to align with the needs of their startup, which they hope to work on full-time after graduation.

While MarHub is now limited to the public Facebook site and the chatbot, the team is building an interactive mobile platform that will enable refugees to view, evaluate, and comment on information from humanitarian agencies, volunteers on the ground, and other asylum seekers.

Nomanbhoy said she sees the company moving from helping people navigate the interview appointment and asylum system to a long-term goal of using data to help improve migration management.

Lee, who teaches Product Design at Berkeley-Haas, praised the team for taking what they learned in class and applying it in Greece. “I think there’s a stereotype about the Bay Area that we’re solving first-world problems for the millennials, but the real issue is the next billion users, and, in some sense, that’s what Sarrah’s team is focused on,” he said. “They really got their hands dirty and went to these refugee camps to get a sense of the real problems for refugees. They’re applying lessons from class to understand those problems.”

— Krysten Crawford contributed to this story.