Vimal Kumar

Dr Vimal Kumar is a postdoctoral researcher at the University of Limerick, where he works on making large language models trustworthy enough for high-stakes use. He is a member of the Biocomputing and Development Systems (BDS) group within Lero, the Irish Software Engineering Research Centre. His current research builds verifiable generation systems for governance documents. Rather than relying on retrieval alone, the framework constrains language models with a domain ontology and a policy knowledge graph, validates each generated document against formal SHACL constraints, and records the evidence, concepts and provenance behind every individual clause — so that generated agreements can be audited rather than simply trusted. Previously at University College Dublin, he developed natural language processing methods for regulatory technology in collaboration with industry partners Corlytics and Version 1, supported by the Irish Department of Enterprise, Trade and Employment through the Disruptive Technologies Innovation Fund (DTIF). The work sought to streamline regulatory compliance while safeguarding consumer interests. Dr Kumar has nearly two decades of research experience and 20 peer-reviewed publications spanning machine translation for low-resource languages, text summarisation, document classification, contextual representation learning and large language model systems. He reviews for ACM Transactions on Asian and Low-Resource Language Information Processing, IEEE Access and Soft Computing, and serves on the programme committees of various conferences.