Madusanka Premaratne
About & Research Leadership

The research behind the work

I'm Founder, CEO & Chief AI Research Officer at Knivok, leading applied research on privacy-preserving LLMs, query-time semantic generalization, and on-device agentic RAG architectures.

My research produces manuscripts accepted for AIRC 2026 and EICON 2026, supported by active university-level collaborations.

I hold a First Class BSc in ICT from the University of Jaffna (GPA 3.93/4.00, ranked 1st in batch) and an MBA with Merit from London Metropolitan University.

Off the clock, I'm a self-confessed petrolhead — happiest behind the wheel of something fast. The Knovik team and I like to ship at speed too, just with fewer traffic violations.

10+
Years in Industry
5
Published & Accepted Papers
Privacy & Edge AI
Research Tracks
95.1%
Semantic Privacy Rate

Core Technical Domains

Large Language ModelsPrivacy-Preserving AISemantic GeneralizationAgentic RAG SystemsEdge AI / On-DeviceNLP & Text IntelligenceAI Product StrategyResearch MethodologyPython · LangChain · RAGDocker · REST API · CI/CD

Research Focus

Query-time semantic generalization, on-device agentic RAG for candidate evaluation, and attribution stability audits for organizational models.

Industry Application

Translating theoretical guarantees into production-ready software — LLM inference pipelines, multi-agent orchestration, and data-sovereign systems.

Recognition

Best Poster Presenter at EICON 2026 for EdgeTal. Accepted Author at AIRC 2026. Invited panelist on digital entrepreneurship on Sri Lanka national television.