Institutional Background
The Department of Artificial Intelligence (DoAI) and the Information and Language Processing Research Lab (ILPRL) collaborate at the School of Engineering, Kathmandu University, to address one of the most critical challenges facing technology in Nepal: the representation, preservation, and computational utility of Nepali and Nepal-based languages in modern Artificial Intelligence.
Department of AI (DoAI)
Offering Nepal's first B.Tech in Artificial Intelligence, focusing on modern deep learning architectures, multimodal AI, data engineering, and ethical AI deployment.
Visit doai.ku.edu.np ↗ILPRL (Established 2004)
Pioneered Nepali Natural Language Processing, Devanagari OCR, Nepali morphological analyzers, the MPP 43-POS tagset, and electronic lexicon archives.
Visit ilprl.ku.edu.np ↗Faculty Mentorship & Direction
The capstone internships indexed on this platform are conceived, steered, and reviewed under the joint academic guidance of faculty leadership:
Mr. Sunil Regmi serves as Lecturer and Bachelor Program Coordinator at the Department of Artificial Intelligence, Kathmandu University. He coordinates undergraduate AI capstone research, ensuring that student implementations satisfy rigorous AI engineering specifications, scalable architecture design, and academic accreditation standards.
Prof. Dr. Bal Krishna Bal is Professor of Computer Engineering, Dean of the School of Engineering, and the Founding Lead of the Information and Language Processing Research Lab (ILPRL) at Kathmandu University. A pioneer in low-resource computational linguistics, he has led foundational research in Devanagari OCR, Nepali WordNet, machine translation, and language resource standardization for over two decades.
Guiding Principles for Student Research
- Strict Low-Resource NLP Focus: Projects must directly address linguistic, data, or modeling barriers in Nepali or Nepal-based languages.
- Auditability & Provenance: Datasets must preserve complete origin, license, and annotator lineages rather than producing opaque web scrapings.
- Public Deliverables: All verified software tools, datasets, and benchmark cards are made available open-source on GitHub and Hugging Face under permissive academic licensing.