Research
My work sits between engineering and research: building systems that make data trustworthy, and applying machine learning to problems where correctness and explainability matter more than raw accuracy.
Those interests run across data-quality engineering, applied machine learning and security, research information systems, explainable medical imaging, and language technology for Afaan Oromoo — a language with limited digital resources.
Areas of Work
Strands of ongoing engineering and experimentation. Each is evidenced by public code — none of these are published papers.
How data-quality expectations can be expressed, checked continuously and surfaced early enough to act on — covering validation rules, anomaly and schema-drift detection, and the workflows that follow a failed check.
Related work: Data Quality Monitoring
Feature extraction from URLs and QR payloads, machine-learning classification of risk, and presenting that assessment to a non-specialist in language they can act on.
Related work: TurkQuish
Segmentation and classification of brain tumours from MRI across several architectures, with attention to what drives a prediction — a prerequisite for any model intended to support clinical judgement.
Related work: Brain Tumor Analysis
Aggregating institutional research into a single discoverable corpus, and the retrieval problem that follows: semantic search over heterogeneous records, and assistance that stays grounded in retrieved evidence.
Related work: ResearchHub Ethiopia
Building usable language tools for Afaan Oromoo, where digital resources are scarce: bilingual lexical search, offline-first delivery for low-connectivity settings, and the data structures that make both practical.
Related work: Afaan Oromoo Dictionary
Status
Konya Technical University