Atif Javed Uses AI to Ensure No One Gets Lost in Translation

Can AI break language barriers? Atif Javed is proving it can.

Atif grew up as an interpreter for his refugee and immigrant family in the U.S. He saw firsthand how language could open doors — or slam them shut. After graduating from MIT during the Syrian refugee crisis, he knew he had to act. Atif swiftly quit his tech job and began building Tarjimly. This AI-powered platform connects refugees and humanitarian workers with bilingual volunteer translators in real time.

Tarjimly uses AI to match, train, and pre-translate languages, enabling faster, more accurate connections between volunteers and those in need. By leveraging deep data on low-resource languages, such as Rohingya and Swahili, Tarjimly fine-tunes large language models (LLMs) specifically for underrepresented languages. This targeted approach addresses the shortcomings of mainstream LLMs, which often lack robust support for these languages.

What started as a bold idea now powers a global network of thousands of translators — many of whom are refugees themselves. Tarjimly’s AI doesn’t just solve a problem. It bridges divides so every person’s story gets told and every need gets met, no matter where they are or what language they speak.

We sat down with Atif to learn how he’s using AI to ensure no one gets lost in translation.

How did your journey inspire you to explore AI for humanity?

Growing up as an immigrant in the United States, I was the primary interpreter for my family. I saw firsthand how critical language access was for navigating day-to-day life, especially for my grandmother at places like my school or the doctor’s office. As an adult, I volunteered in refugee camps in Greece and Turkey where I witnessed how language barriers prevented people from accessing essential services. It became clear that communication gaps were a severe issue, not just a minor inconvenience.

At that time, my friends and I worked in the tech industry. We saw AI’s untapped potential to meet the urgent needs of refugees and immigrants and the opportunity for it to transform the humanitarian sector. So we launched Tarjimly. It’s a mobile app that connects refugees or humanitarians with bilingual volunteer translators in real time by leveraging AI for matching, training, and even pre-translation — where AI translates parts of conversations before a translator steps in. AI enables us to scale this impact, ensuring faster and more accurate matches between translator and refugee, while amplifying the human empathy at the heart of every interaction. Tarjimly ensures that no one remains unheard or unseen — because I believe it’s a human right to be understood.

How do you ensure that AI tools are developed responsibly, especially when working with vulnerable communities like refugees?

Developing AI tools responsibly begins with a deep understanding of the communities we serve. At Tarjimly, we prioritize co-creation with refugees, immigrants, and humanitarians, ensuring their voices shape every aspect of our tools. Our community of over 61K volunteer translators includes immigrants and refugees with firsthand experience navigating language barriers. Their lived expertise makes them uniquely equipped to deliver meaningful solutions. Their insights guide the design, functionality, and refinement of our AI tools.

For example, focus groups with our refugee beneficiaries revealed that many refugee women prefer women translators. In response, we created a women translator filter to meet this need. Similarly, feedback from fieldworkers and refugees highlighted the necessity of asynchronous translation options to accommodate unpredictable schedules, prompting us to implement this feature.

Our AI and machine translation features are validated and improved by our translator community. Human translators directly verify AI-generated translations, provide feedback, and ensure linguistic and cultural accuracy. This work makes the tech better with every iteration. By placing refugees and immigrants at the heart of our development process, we empower those most affected by displacement and other humanitarian crises to lead the way in overcoming shared challenges, as well as fostering ownership and trust in the tools we build.

In what ways do you think AI will transform the landscape of humanitarian aid over the next few years?

AI has immense potential to revolutionize humanitarian aid. We’ve already seen incredible progress in issue areas like language access. When we launched our generative AI feature, FirstPass, we were amazed by how much generative AI pre-translations improved the accuracy of final outputs. This breakthrough was made possible through our participation in Google.org’s inaugural GenAI Accelerator, which played a pivotal role in shaping and advancing this innovation. With advancements like instant voice language detection and translation, the pace of innovation is accelerating. I envision a future where AI becomes reliable even for sensitive use cases, providing real-time, accurate support to those in need.

Beyond language access, AI is reshaping the broader humanitarian landscape. AI can unlock aid faster, making it more efficient and more accessible than ever before. In my participation at conferences and summits, I’ve seen my peers leverage AI to inform early detection systems, crisis prevention, and sustainable resilience strategies. These tools can predict disasters, monitor risks, and build preparedness before a crisis strikes. The next few years will be transformative as we expand from crisis response to proactive solutions for resilience. I believe AI will unlock new possibilities, helping us reach more people and create lasting, impactful solutions for vulnerable communities worldwide.

What core values drive your unique vision for impact in an AI-driven future?

It’s simple: AI should exist to help people — all people, and especially those who need it the most. Too often, technological advancements benefit a privileged few, from a certain geography or language community, leaving historically marginalized communities overlooked. My mission is to ensure that AI becomes a tool for equity, breaking down barriers rather than reinforcing them.

I believe in prioritizing the underserved — refugees, immigrants, and those living in crisis zones — because they are the ones who often face the greatest challenges yet receive the least attention when it comes to innovation. Even when resources are scarce or investment priorities lie elsewhere, my commitment is to create AI tools that uplift and amplify the voices of those who have been neglected for far too long. Success shouldn’t just be measured by technological breakthroughs but by the impact we make in bridging divides and fostering inclusion.

Which visionary leaders, philosophies, or movements give you hope for a more human-centered AI future?

In working with funders, technologists, and social entrepreneurs, I’ve noticed a growing realization: while the digital boom brought remarkable progress, it also widened the digital divide, leaving many without access to its benefits. Now, as AI reshapes our world, there’s a shared understanding that we must actively prevent an AI divide from becoming a byproduct of this revolution. Instead, we need to ensure that AI benefits everyone, not just those with the most resources. This commitment to equity and inclusivity inspires me and gives me confidence in the future of human-centered AI.

At the grassroots level, I’m deeply inspired by our refugee translators who, even while living in camps, are stepping up as leaders and drivers of change. They are not only helping their communities overcome critical language barriers but also collaborating with us to shape and refine technology that directly addresses their needs. Their resilience and commitment to leveraging technology for good remind me of the true potential of AI.

What is your 7-word autobiography?

Humanitarian heart, technologist mind, and changemaker spirit.