At Ethiopian Artificial Intelligence Institute, I design, train, and evaluate computer vision and applied AI models for healthcare-relevant use cases. I lead AI researchers and engineers through the model lifecycle, including developing lightweight models for low-resource deployment with knowledge distillation.
I built a Retrieval-Augmented Generation pipeline for a blended learning and tutoring platform, engineering retrieval over course content for learner-facing question answering. I also developed an internal LLM-based conversational agent and co-authored computer vision and applied AI papers accepted in journals including Scientific Reports.
Alongside this work, I develop React and Django REST Framework applications that serve AI/ML inference endpoints, and build ELT pipelines and monitoring dashboards. My projects also include healthcare models such as Deep Eye, a knowledge-distilled MobileNet classifier for diabetic retinopathy staging whose results were published in Scientific Reports.
