Tarjimly: On-demand Translation for Displaced People

Refugees and immigrants severely lack language support during resettlement, such as asylum or housing applications, understanding laws and voting, seeking employment, education, healthcare, or simply going about daily life. Our solution is a mobile app that mobilizes the world’s 3 billion bilinguals as on-demand translators for the 68 million displaced people and the humanitarian organizations that support them. There are potentially 120 connected bilinguals for every 1 connected refugee. We use machine learning matching, training, and calling tools to connect the best available volunteer in the world to someone requesting language assistance in just 60 seconds. We mobilize the remote skills of our 10,000-strong community, empower refugees with language access and employment as translators, and improve the efficiency and access of humanitarian support. Instead of building walls of misunderstanding, we give megaphones to the voiceless, not only to be heard, but to thrive in a connected world.

The problem is that translation services are prohibitively scarce and costly for refugees and immigrants. The world is seeing an explosion of forced displacement expected to increase to 143 million by 2050 due to climate change and conflict. People crossing borders and continents don’t speak the native languages of their host countries creating major hurdles to social inclusion, especially for non-English speakers, and massive inefficiency for service providers. Refugees and service providers in camps or resettlement settings wrestle with language support because (1) the demand for translation far outweighs the local supply but not the global, (2) refugees are often at the mercy of their providers and disempowered in learning new languages to drive inclusion and employability, (3) machine translation isn’t used in the field because it’s not robust enough yet for the long-tail of languages and dialects and in riskier settings, and (4) providers are forced into enterprise fees or to manage their own low-tech translation department. Language is one of the biggest risks to service providers providing effective support - a doctor in a refugee camp can serve 3x as many people and improve treatment fidelity simply by having a quality translator available.

Our mission is to improve the lives of refugees and immigrants by eliminating language barriers with the world’s most accessible translator service. Refugees are trapped not just by borders, but language barriers that prevent access to quality interpreters, so they often rely on children, receive worse service, or flat-out can’t be serviced. Today, most have smartphone and internet access as a lifeline to their families, but they’re jaded by inaccurate machine translation and hobbyist techies. We’ve spent 3 years co-designing our solution with refugees in camps in Greece and Turkey interviewing over 300 refugees and aid workers, connecting 20,000+ of them to a human, often from their diaspora, in an average of 60 seconds with chat, VoIP calls, and streamlined user feedback. Social workers began using our app to help with resettlement, and we now have running pilots with field partners in multiple states and countries. Technology must work with communities to build trust. Over 90% of our team are from refugee or immigrant families - we were always there when they needed translation so we’ve seen how deeply it impacts them. We’re willing to make it our life’s work to mobilize the whole world to do the same.

Our solution merges the traditional community-model with first-class technology:

1. We provide a free Messenger app that makes it easy for any bilingual to signup and train to become a translator and for any refugee or immigrant in need to anonymously and immediately access a translator in over 100 languages. We use Twilio APIs to create a free calling solution for refugees in over 20 countries no matter where their translator is located.

2. We provide a freemium Mobile App targeting humanitarian workers and service providers who need a higher range of features, translator quality, and confidentiality. We use machine learning models to score volunteers, eliminate scheduling, and guarantee a translator in 60 seconds. We plan to use our massive dataset to help with refugee needs assessments and train NLP models that can further support scaling demand.

3. Language isn’t always a barrier, rather, it can also be one of the best entry points for employability and social inclusion for refugees or immigrants who become qualified translators and interpreters. Our platform also creates empowerment by training, certifying, and hiring refugee translators to service the humanitarian organizations in need of premium translators. This gives refugees a pathway to inclusive economic growth and greater access to opportunities with fair pay.

Tarjimly innovates at both a product and human level by mobilizing thousands to easily help displaced people with a skill they already have: language. Our product uses machine learning matching algorithms to predict availability and quality for any pool of global volunteers at any given time. This props up our unique model of expanding the global pie of translators and interpreters available by training and enabling every bilingual to help. We built a better quality model than industry standards by using a multitude of factors, including response times, ratings, and experience. Translator management used to be prohibitively expensive, opaque, and slow, but now it's just a tap away, especially on messaging channels and international VoIP calls which displaced people heavily rely on.

Tarjimly also lowers the commitment bar for volunteers, raises the certainty of impact, and deepens connection, empathy, and understanding. This all leads to a massive community of people eager to do something more than donate and post online to help displaced people. The sheer size of the community allows unprecedented routing such as specific dialects, unique expertise, or even gender and LGBTQ-friendliness. 70% of our translators are women and having a female translator can make a world of difference for an Iraqi woman who has faced miscarriage, sexual abuse, and depression. This unprecedented language access has the potential to improve the quality of life for those struggling to find a home in the world and even save lives.

Machine learning matching algorithms and mobile app calling technology.

When a refugee or aid worker requests a translator on Tarjimly, our algorithms that have trained for 2 years on over 70,000 users, determine which translators are best suited to ping. This reduces the number of people we bother and reduces the time it takes to connect to a translator from 5 minutes to 2 minutes down to our current average of just 57 seconds. If Tarjimly simply bothered thousands of people simultaneously at random times it would be just like every other online spreadsheet or group that calls upon everyone to see who can help. When translators are pinged on Tarjimly, they have a high confidence that they are the right person and will be chosen to impact someone's life. Our algorithms have make deploying and finding the right volunteers more efficient.

The Tarjimly mobile app also implements innovative calling technologies to connect people anywhere in the world. With the help of our technology partner, Twilio, regardless of where a refugee is and which mobile device they have, they can be connected to a translator through either a VoIP call or a fallback GSM conference line at no cost to them. We operate Twilio numbers in multiple countries that route together a conference call, for example, with a refugee in Greece and a translator in Canada - and even add an attorney in the UK to the call as well! This creates unprecedented access to a global pool of support.

We conducted surveys of 80 aid workers in Greece with varying lines of work (case workers, coordinators, teachers, etc.) which showed that having a quality remote interpreter immediately available could double their efficiency when supporting refugee communities. A similar survey of 8 medics in Greece and Turkey consistently showed that it would take 30 minutes to diagnose and treat a patient without a translator versus just 10 minutes with. One of our solution partners is a United Nations agency that is helping us run a pilot of Tarjimly in a refugee camp to measure the increased efficiency.

We publish our real-time impact metrics at https://tarjimly.org/data

Currently, we are serving almost 20,000 registered beneficiaries (refugees and aid workers that proxy for refugees) with our 10,000+ volunteer translators.

One Year: 100,000 beneficiaries, 50,000 volunteers

Five Years: 1 million beneficiaries, 5 million volunteers

Our goal in the next year is to unblock critical language barriers by providing translators for 100,000 displaced people, and for 1 million within 5 years.

Our goal is to be a trusted tool in the refugee and humanitarian communities that gives hope. Today we are the "disaster" solution designed to provide access when no other options exist, but in 5 years we want to be the default solution that millions of displaced people go to when they need language support in over 10 countries. In 1 year, we want to have 5 large humanitarian partners on the ground that we are servicing with hundreds of translations weekly.

As an organization, we want to grow to 5 people by next year and even in 5 years, we don't want to grow to beyond 10-12 people. We want to only work on highly scalable solutions and not be beholden to fundraising to grow, instead be a self-sustaining nonprofit.

Our biggest barrier is the expertise to convert humanitarian/NGO customers. We know our service can outperform the status quo and the market is massive and growing, but closing the sales in a reasonable timeframe with the right legal and service-level backing is going to be the biggest challenge facing us alongside staying true to serving the refugees.

In 5 years, I'm confident we will find the sweet spot for our service that will make demand grow exponentially. At that point, despite having a small team, the financial barriers of supporting a widely-used service with thousands of sessions a week without debt financing will be challenging, but achievable with the right success in our sales and revenue-sharing.

Patience and persistence. We will overcome these barriers by staying open to iteration when something isn't working and keeping lean as a team. We are interviewing and confident we'll bring the right people on the team to support these current gaps. Sincerity and putting the refugees' needs first will benefit our solution greatly as we try to earn trust and respect in a relatively small space of humanitarian organizations. We plan to leverage our board and our existing finances to setup for this long-term plan.

N/A

4 full-time staff (Executive Director, CTO, Software Engineer, Community Manager). The rest of our team is composed of 15 passionate volunteers around the world who support engineering, data science, marketing, and PR.

My grandmother was a refugee who fled partition in 1947. My family was scheduled to immigrate to the US the week after 9/11. When we finally made it to DC, the fear of brown-skinned Muslim immigrants was palpable. But I was luckier than my grandmother who used to rely on me as her daily interpreter. She would tell me "at least people judge you after they hear you, instead of just after they see you, like me". Language was the crutch I used to pull myself to MIT and Silicon Valley alongside my best friend and roommate of 7 years, Aziz. We felt helpless watching Syrian mothers being demonized daily with no voice of their own or support around them.

My team and I are the children of refugees and immigrants. We have a responsibility to use our skills to help people like our parents and preserve our mother tongues. We have MIT and Stanford engineering degrees, complimentary personalities and skillsets, and some of Silicon Valley's strongest tech companies on our resume. But the refugee crisis keeps us up at night and gets us up in the morning and that's why we quit our jobs to pursue Tarjimly full-time.

We are working with several technology partners to perfect our product and market our service providing engineering support, credits, and mentorship: Twilio, AWS, Facebook, Google. We also partner with CSR at these companies and others like BlackRock to provide employee volunteering programs on our platform and systematically grow the volunteer pool.

We also are closely partnering with humanitarian organizations to deliver our solution: the United Nations Capital Development Fund, Boat Refugee Foundation, Translators Without Borders, Refugee.info, United Work.

Our business model has two components:

1) The first is to sell a premium translation service to humanitarian organizations and governments that are heavily involved with immigrant communities. We offer paid features, preferences, and access to Tarjimly certified translators. This would partially cover our costs towards the free refugee tool and still largely benefit displaced people via aid workers, including immigration attorneys, medics, teachers, NGO staff, etc.

2) The second is the refugee paid translator program which gives translation credentials and paid opportunities to qualified bilingual refugees whom we revenue-share with. This would cover the other portion of costs towards our free refugee tool and provide a path to economic empowerment for displaced people.

Our path to financial sustainability is outlined in detail in the "Revenue" tab here: https://docs.google.com/spread...

We plan to transition during 2020 and 2021 from primarily foundation and accelerator grants to a mix of:

1) translation service customers (whether they're NGO/government or even regular business for our refugee translator program) and

2) small individual donations across a massive pool of volunteers.

The first would be the Employment model. Both sources would fit the Service Subsidization model since both revenues and and Employment models. Our objective is to be self-sustaining by 2023 (5 years).

We believe the Solve network can open doors for us much faster than by ourselves, especially humanitarian-centered technologists. The funding from Solve would be monumental for our team to execute at full pace on our revenue model. The MIT community was the most invigorating and inspirational group of people we'd ever been with and we hope we can continue meeting and bounce ideas off such incredible people through Solve. 

We are also looking for board members and advisory members with unique skills to complement our team. We built a strong process of bringing on volunteer engineers with well-defined projects for them to work on a part-time basis, but we are in eager need for someone with business development and sales experience to help execute our team's our revenue strategy now that we have launched our mobile app. Lastly, Solve financial support is always valuable as we seek to grow our org now that we've mastered our pitch and story. We also can use champions within companies to grow our employee volunteering program and volunteer marketing.

N/A

We are eager to partner with international humanitarian organizations, especially those that support resettlement and medical aid.  We are eager to partner with the Red Cross, The International Rescue Committee, CARE, Doctors Without Borders, Relief International, Oxfam to name a small few as well as private sector businesses that rely heavily on translation for international work. That is also why we are actively working on fundraising and hiring for a business development team member or board member who can structure good partnerships and service agreements with these organizations.

For larger organizations with an existing translator workflow, Tarjimly can also leverage their existing translators and training. For example, with Tarjimly an IRC specialist in Sacramento could access an IRC volunteer translator in the NYC office with a fallback of Tarjimly certified translators followed by regular Tarjimly volunteers. Generally, we would like to partner as technology experts where we have an open and regular feedback loop with NGOs.

We would utilize the funds from this prize in our refugee paid translator program so we could go from 10 to 500 refugee translators and interpreters. This would open doors to the paid translation opportunities in the private sector, law offices, hospitals, and more and it would allow us to build the right training and certifying modules on Tarjimly so that these refugees can become self-reliant on their own careers. Our objective is to create a pathway to economic empowerment for refugees, and without the language barrier, we can ultimately highlight how much refugees contribute to society when they are given the chance to thrive. This prize would also support our team creating showcase studies of how language access promotes inclusivity and belonging for refugees in their new communities. These funds would help us increase both quality and access to our language support services for thousands of refugees around the world.

Right now we don't have enough funding for all this, but winning this prize would take us across the finish line and prove this idea can scale!

We would utilize the funds from this prize in our refugee paid translator program so we could go from 10 to 500 refugee translators and interpreters. This would open doors to the paid translation opportunities in the private sector, law offices, hospitals, and more and it would allow us to build the right training and certifying modules on Tarjimly so that these refugees can become self-reliant on their own careers. Our objective is to create a pathway to economic empowerment for refugees, and without the language barrier, we can ultimately highlight how much refugees contribute to society when they are given the chance to thrive. This prize would also support our team creating showcase studies of how language access promotes inclusivity and belonging for refugees in their new communities. These funds would help us increase both quality and access to our language support services for thousands of refugees around the world.

Right now we don't have enough funding for all this, but winning this prize would take us across the finish line and prove this idea can scale!

We would utilize the funds from this prize in our refugee paid translator program so we could go from 10 to 500 refugee translators and interpreters. This would open doors to the paid translation opportunities in the private sector, law offices, hospitals, and more and it would allow us to build the right training and certifying modules on Tarjimly so that these refugees can become self-reliant on their own careers. Our objective is to create a pathway to economic empowerment for refugees, and without the language barrier, we can ultimately highlight how much refugees contribute to society when they are given the chance to thrive. This prize would also support our team creating showcase studies of how language access promotes inclusivity and belonging for refugees in their new communities. These funds would help us increase both quality and access to our language support services for thousands of refugees around the world.

Right now we don't have enough funding for all this, but winning this prize would take us across the finish line and prove this idea can scale!