The Problem

I have been working as a volunteer helping marginalized communities for several years now. 3 years ago I started offering STEM after-school programs for refugee kids, thinking that it would significantly improve their future career outcomes. I was taken aback however, when I realized that 80+% of the kids couldn’t take advantage of my program due to their lack of English proficiency. I decided to pivot my program as a result and started to look into the languages they spoke and the extent of the problem.  I learned that most languages they spoke were “low-resource” languages (i.e. dialects or languages with less data) such as Somali, Amharic etc with little support from major translation companies. I also learned that there are 5.1 Million English language learners (ELL) in the US with 50k Refugees coming into the US each year! Many low resource language speakers however do not receive adequate support from our education system even though research shows that taking advantage of one’s native language helps accelerate language learning. The root cause is the lack of  multilingual teaching material and staff. 

I wanted to provide a more scalable solution to this problem than unsuccessfully recruiting multilingual teachers for the 200+ languages spoken in King County!  I also wanted a solution that not only taught English but also helped preserve language and cultural diversity. 

Your Solution

I researched and discovered that existing products for language translation such as Duolingo or Rosetta Stone focus only on a dozen HIGH resource languages, not languages and dialects that refugees speak and do not offer live auto translation feature that preserves the original text in the book. Also, their content is not culturally sensitive. I therefore decided to come up with ideas to build fun and engaging bilingual content apps for LOW resource language speakers to learn English. I established a nonprofit organization, Linguistics Justice League (LJL), and hired over 40 volunteers to focus on this effort. Through my nonprofit I created a multilingual library app called EduLang. The app leverages Machine Translation APIs that focus on low resource language translation and translates children’s books into 108 languages. Users can also upload their own books and the app can scan the upload, extract and translate the text in the book into the user’s desired language and present the translated text next to the original in the book. So far we were able to serve 12+ organizations and over 250 learners. We conducted user surveys to build our roadmap. Additionally, content is always accessible in both English as well as the learner's native language unlike other industry translation applications, thus leveraging and preserving the learner’s proficiency in their native language.

Founding Story

I offered STEM after school programs at refugee organizations to female refugee youth. I initially noticed that less than half of the students attended the STEM programs and I assumed it was due to a lack of interest. However, after talking to the director, I realized that many refugee youth were passionate about learning STEM but lacked English language skills. This led to an "aha!" moment for me! I then decided to shift my focus to apply technology to help with English learning at scale.  I learned an important lesson to start from the needs of the person I am serving to be more effective.

Your Impact

I am proud of the way the technology my team developed has served the community and helped students learn. Hundreds of learners using our EduLang app have been able to read bilingual books across 108 languages such as Swahili, Ukrainian, Somali etc. The app won the US Congressional App Challenge. Despite the progress thus far, there is still more to do!  We received a request from an organization in Ghana to enable the Twi language for example. 
We want to take this app to the next level by increasing the number of languages we support and by focusing on speaking, listening and writing. We plan to add an assessment to identify learners’ speaking and writing levels and personalize content.  We plan to provide the ability for the learner to hear stories from the library and read them aloud while the app provides feedback by leveraging text to speech and speech to text technology.  For writing proficiency we plan to add fun games such as Mad Libs with multilingual prompts. Additionally,  we plan to source better content by partnering with a language translation organization and a children's content creator. We want to recruit additional volunteers and increase our impact from a few hundred to tens of thousands of users in the next 12 months by raising awareness of the app. We are barely getting started on enabling linguistic justice for the thousands of languages spoken worldwide!

Video