My engagement with data curation and translation challenges drew my attention to the linguistic and technological gaps shaping everyday translation practices. Over time, I observed that many translators operate within emerging gig-based, platform-mediated labor systems—highlighting a critical intersection between language work and digital employment. These insights shaped my research focus on the relationships among technology, language, labor, and translation. Therefore, I decided to focus my research on Kenya. These experiences ultimately shaped the trajectory of my dissertation, Translators as Gig Workers: Re-iterating White Collar Work in the Era of Digital De-Professionalization in Kenya”, which broadly explores how digital platforms are transforming the nature of gig work, particularly translation work in Kenya.
My research examines how platform-mediated systems are restructuring translators’ employment models and redefining professional identity within the country’s growing gig economy. I position translators as central actors in this ecosystem, analyzing how they navigate the uncertainties of digital labor while also probing issues in translation technologies (e.g., Computer-Assisted Translation tools) and the role of generative artificial intelligence in translation, particularly in low-resourced languages. Finally, I emphasize the importance of language, particularly African languages that remain underrepresented in global AI systems. Building on this commitment, I extend my work through an applied project focused on curating language data and developing resources for multiple Swahili dialects. This effort addresses a critical gap in computational linguistics by foregrounding cultural and linguistic diversity in dataset design.
Drawing on fieldwork in Kenya, the core chapters of my dissertation explore how gig translators navigate structural components and digital tools in their everyday work. Dorothea Kleine’s Choice Framework, informed by Amartya Sen’s Capabilities Approach, provides the theoretical grounding, highlighting how workers’ well-being depends on their ability to exercise genuine agency within unequal digital labor environments.
Freelance translators in Kenya operate in a gig-based digital ecosystem. They juggle multiple platforms, clients, and deadlines while managing uncertain income and limited professional recognition. In the first study, I interviewed translators to understand their resource portfolios—a combination of skills, education, networks, infrastructure, and social support that enable them to survive and thrive in this environment.
Insight: Translators rely on diverse resources to navigate instability, highlighting the intersection of human capability and digital labor.
Next, I explored the tools translators use, particularly Computer-Assisted Translation (CAT) systems. While these tools offer efficiencies such as translation memories and term databases, they often fail to capture local languages, idioms, and cultural nuances, limiting their usefulness in African contexts.
Insight: Tool design determines not just translation quality but also professional opportunities—translators whose expertise is marginalized face fewer career options.
Finally, I examined Generative AI (GenAI) and its role in translation. Freelance translators in Kenya increasingly experiment with tools like ChatGPT. My experimental study revealed a mixed landscape: Although AI often produces correct translations, it frequently struggles with idioms, contextual interpretation, and nuanced pragmatic meaning. For instance:
English: Ring also settled a lawsuit with a competing security company, the ADT Corporation.
Human: Ring pia alisuluhisha shtaka na kampuni ya usalama la muungano wa ADT
AI: Ring pia alifikia makubaliano ya kisheria na kampuni inayoshindana ya usalama, ADT Corporation.
Observation: The entire sentence was translated directly, without taking the context into consideration.
Insight: Generative AI cannot yet replace human translators. Instead, it creates new opportunities for skilled post-editing labor, where human expertise ensures translations remain authentic.
Mapping the intersections of freelance/gig translators, translation technologies, and AI has direct, practical implications for the applied world. By analyzing translators’ resource portfolios, organizations can identify gaps in education, access to technology, and professional networks, enabling targeted interventions such as training programs, mentorship, or infrastructure support to improve income stability and career growth. Insights from CAT tool analysis reveal limitations in supporting local languages and capturing cultural nuance, providing a clear roadmap for developers to design inclusive, context-aware translation technologies that better serve low-resource language communities. Similarly, examining human-AI collaboration highlights where skilled post-editing is required, informing the creation of efficient workflows that combine AI productivity with human expertise. Collectively, these mappings enable policymakers, NGOs, and technology developers to make data-driven decisions that strengthen digital labor ecosystems, reduce socioeconomic disparities, and ensure that AI tools reflect linguistic and cultural realities, creating a more equitable and practical landscape for translators and technology users alike.