Leonard Jombo
Computer Science · Research . & NLP
I'm Leonard Jombo, a Computer Science graduate from Maseno University. I build backend systems, mobile apps and developer tools in Go, Python and Flutter. I also spend time on NLP research: in 2026 I took part in the Princeton in Kenya Global Seminar, where I built TaitaCLIR, a cross-lingual search project for the Taita language and got a firsthand look at how embedding models perform on low resource languages.
Background
I hold a BSc in Computer Science from Maseno University (2022–2026). My interest in technology began with curiosity about how software breaks, which led me to cybersecurity. Over time, that curiosity grew into a broader interest in how reliable systems are designed and built.
Most of my hands-on experience is in backend and mobile engineering. I've built REST APIs, WebSocket services and relational database systems using Go. My projects include Dawida, a mobile app for the community-based curation of Taita language datasets(text and speech); Mawora, a farm management app and Kairo, a real-time project collaboration platform. My security background has also shaped how I approach software development, with an emphasis on correctness, reliability and building secure systems.
I've also gained research experience through the Princeton in Kenya Global Seminar, where I worked on TaitaCLIR, a cross-lingual information retrieval project for Taita. That experience exposed me to the shortage of structured data and language technology tools for low-resource African languages. Addressing that gap is a key reason I'm increasingly drawn to machine learning and natural language processing.
Research Interests
My research interest is NLP and machine learning for low resource African languages, especially cross-lingual transfer and community-driven data curation. Most of this comes from working on Kidaw'ida (Taita), an agglutinative Bantu language that shows where a lot of standard NLP tooling struggles.
Dawida taught me that low resource NLP starts with data: capturing words, audio and cultural context directly from speakers before any modelling is possible. TaitaCLIR gave me hands on experience with cross-lingual sentence embeddings and semantic retrieval over a small parallel corpus. I'm now curious about morphologically aware machine translation for languages like Taita, including how similarity to Kiswahili and lightweight fine-tuning methods could make translation models practical for communities with limited digital resources.
Education
Bachelor of Science in Computer Science
Maseno University • 2022 – 2026
Coursework spanned software development, distributed systems, cloud computing and artificial intelligence, with a strong foundation in cybersecurity and information security management. I served as Cyber Security Lead in the Google Developer Student Club, and my interests center on backend systems, secure software design and, increasingly, machine learning and NLP for low resource languages.
