Ndege Ezra

Software and systems architect

Ndege Ezra

Biography

Ezra Ndege is an undergraduate student at Maseno University pursuing a degree in Mathematics and Computer Science. He has deep expertise in software design and development, and his mathematical training shapes how he approaches system architecture, data integrity, and problem decomposition. He contributes to the Maseno Centre for Applied Artificial Intelligence (MCAAI), a university centre of excellence that develops community-driven AI for health, agriculture, education, climate, and governance. The centre is known for its work on speech and sign language data for under-resourced Kenyan languages.


Within MCAAI's innovation portfolio, Ezra has contributed to the foundations of two language technology products. SmartVoice is a community voice data platform for Kikuyu and Kalenjin. It supports contributor onboarding, consent capture, voice qualification, and separate recording paths for speech recognition (ASR) and text-to-speech (TTS). Recordings pass an audio quality gate that checks duration, noise, edge silence, and signal-to-noise ratio. Reviewers and language leads then validate them through single and bulk listening workflows, with role-based governance. Tafsiri is a multilingual localisation system for keywords, sentences, and official project dictionaries. Translators work in exclusive 300-item batches, and proofreaders check a sample of each batch. Approved work feeds translation memory and a versioned term base, and the system produces delivery exports and contributor performance reports.


Both systems are built on modern web technology, including Next.js, Supabase Postgres and Storage, and role-aware authorisation. This lets Ezra's work turn research needs into dependable tools that people use. His contributions help MCAAI build inclusive, high-quality datasets for languages that mainstream technology has overlooked, in line with the centre's commitment to community ownership and ethical data practices. He combines academic rigour with practical engineering and aims to build technology that serves African language communities.


Research Focus

NLP for low resourced languages Computer visions

Publications

No publications on record yet.

Projects

No projects on record yet.

Collaborate with Ndege

Interested in this research? Reach out directly or explore partnership opportunities with MCAAI.

Send Email
All Team Members

Subscribe to our Newsletter