Research · Build · Deploy

Madusanka
Premaratne

Applied AI Researcher

Building AI systems grounded in research — privacy-preserving LLMs, semantic generalization, and on-device agentic architectures.

Start a Conversation View Research

Scroll

The Research behind the work

I'm a Founder, CEO, and applied AI researcher with over a decade in enterprise software and AI systems. I currently lead the research agenda at Knivok Private Limited, focusing on privacy-preserving LLMs, semantic generalization at query-time, and on-device agentic architectures.

My research produces manuscripts in preparation for MDPI and EICon venues, with active university-level research collaborations supporting the work.

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. The work I offer is built on that foundation — not trends.

10+

Years in Industry

Active Research Projects

University Collaborations

95.1%

Privacy Preservation Rate

Core Expertise

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

Research Focus

Privacy-preserving query-time frameworks, semantic generalization for educational LLMs, on-device agentic RAG for talent systems.

Industry Application

Translating research into production-grade AI systems — LLM inference pipelines, multi-agent orchestration, and edge-first deployments for real clients.

Recognition

Selected Lecturer Candidate, University of Sri Jayewardenepura (2025, offer declined). Invited panelist on digital entrepreneurship — national television, Sri Lanka.

What I Build

Services grounded in active research — not packaged templates. Every engagement starts with understanding the real problem.

Research-backed

Privacy-Preserving LLM Systems

Design and implement query-time privacy frameworks for LLM applications — semantic generalization, data minimization, and GDPR-compliant AI pipelines without sacrificing utility.

Production-grade

Agentic RAG Architectures

Build multi-agent retrieval-augmented generation systems for enterprise use — from architecture design through production deployment, including on-device and edge-first configurations.

Custom-built

NLP & Semantic Intelligence

Apply semantic generalization, embedding strategies, and NLP pipelines to solve real classification, extraction, and search problems — grounded in current research, not off-the-shelf demos.

Privacy-first

Edge AI & On-Device Systems

Move inference off the cloud — design on-device agentic systems that preserve data sovereignty, reduce latency, and achieve inherent compliance. No third-party data exposure.

Strategic

AI Research & Strategy

Define your AI research agenda, evaluate model capabilities, and translate academic findings into product strategy. Research collaborations, manuscript preparation, and AI readiness assessments.

End-to-end

Enterprise AI Integration

Integrate LLM inference pipelines and multi-agent systems into existing products — from API design and prompt engineering through evaluation, monitoring, and responsible deployment.

Active Research

Ongoing and completed research projects spanning privacy, NLP, and applied AI — in collaboration with partner research institutions.

2025 – 2026

SemanticGuard

Privacy-Preserving Query-Time Framework for Educational LLMs

Manuscript in preparation — MDPI / EICon

Chief Researcher — Knivok Private Limited

A hybrid agentic framework implementing semantic generalization at query-time to protect educational data prior to cloud LLM submission. Evaluated on 32,593 queries from the OULAD student dataset.

95.1%

Institutional data protection

93.4%

Utility preservation

58–76%

Behavioral fields unmodified by existing systems (gap addressed)

2025 – 2026

EdgeTal

On-Device Agentic RAG for Talent Discovery

Manuscript in preparationVisit site

Lead Researcher — Knivok Private Limited

An on-device agentic RAG system enabling privacy-preserving talent discovery without cloud inference. Addresses data sovereignty in HR technology through edge-first AI architecture — inherent GDPR compliance by design.

Zero

Candidate data transmitted to external servers

Edge-first

Architecture — no cloud inference dependency

GDPR

Inherent compliance by design

2025

Gen Z Job Commitment Study

Determinants of Job Commitment in IT Organisations in Sri Lanka

London Metropolitan University via ESU, Sri Lanka

Principal Researcher — MBA Research Project, London Metropolitan University

Quantitative study with 386 respondents from major Sri Lankan IT firms (WSO2, IFS R&D, Knivok, Prime One Global). SPSS-based statistical methods — correlation and chi-square analysis — to identify key drivers of Gen Z job commitment.

386

Survey respondents

SPSS

Statistical analysis methodology

Merit

MBA awarded with Merit

2024 – 2025

EV Adoption Forecasting

Deep Learning-Based Market Prediction for EV Adoption Among Sri Lankan Gen Z

Applied research — Knivok Private Limited

Research Lead — Knivok Private Limited

AI-driven demand forecasting model integrating primary survey data. Developed predictive insights for EV adoption policy in Sri Lanka, combining deep learning with qualitative market analysis.

Deep Learning

Demand forecasting model

Primary Data

Survey-integrated methodology

Policy

Strategic insights for EV adoption

Professional Journey

From enterprise software engineering to founding and leading an AI research company.

Founder, CEO & Chief AI Research Officer

Knivok Private Limited

Mar 2022 – Present

Founded Knivok with a vision for AI-driven technology solutions. Currently leads the company's applied AI research agenda — directing work on privacy-preserving LLM systems, semantic generalization, and on-device agentic RAG architectures.

Active research collaborations with partner universities and ESU, Sri Lanka
Led development and launch of Cresuma — flagship product contributing 25% of annual revenue
Driving LLM inference pipeline integration, multi-agent systems, and edge AI deployments
Intern and trainee pipelines with University of Vavuniya and University of Sri Jayewardenepura
30% revenue growth in first year; 100% remote team with 40% productivity improvement

Co-Founder & CIO

Insfra Technologies (Pvt) Ltd

Jan 2016 – Feb 2022

Co-founded Insfra Technologies, the first IT company in Badulla — addressing a critical regional gap in advanced technology services across an underserved area of Sri Lanka.

Led cross-functional teams across product delivery, infrastructure, and client engagement
Drove local economic development through employment creation and structured skill programs
Improved project efficiency in a geographically challenging environment

Software Engineer (Associate)

WSO2 Lanka (Pvt) Ltd

Jul 2014 – Dec 2015

Enhanced WSO2 API Manager functionality across enterprise deployments. Maintained product integrity across release cycles and shared technical expertise globally through webinars.

Ensured seamless version migrations and backward compatibility across enterprise deployments
Resolved escalated customer issues, reinforcing platform reliability
Increased product visibility and community engagement through global knowledge sharing

Community & Leadership

Building Sri Lanka's digital economy through community, mentorship, and public advocacy — alongside the technical work.

Executive Director & Mentor

Online Entrepreneurs Club of Sri Lanka (OECSL)

Jan 2017 – Dec 2022

Led workshops, webinars, and mentorship programs equipping Sri Lankan youth with digital business and entrepreneurial skills. Built and scaled a national community of entrepreneurs focused on knowledge-sharing and digital economic growth.

Grew community from 20,000 to 150,000+ members across Sri Lanka
Organised national-level digital entrepreneurship events in collaboration with government bodies
Partnered with the Galle District Digital Transformation Committee on the digital economy initiative

Mentor

Uva Startup Hub

Feb 2016 – Dec 2022

Mentored early-stage freelancers and entrepreneurs in Uva Province — an underserved region of Sri Lanka — supporting their development and contribution to the country's emerging digital economy.

Guided founders through early-stage product and business model development
Contributed to local employment creation and digital skill development programs

Media & Featured Appearances

National television and public events as a thought leader on digital entrepreneurship.

TV Derana2022

Aluth Parlimenthuwa (New Parliament)

Invited panelist representing Sri Lanka's digital entrepreneurship community. Discussed the future of internet entrepreneurship and the digital economy on national prime-time television.

Sri Lanka Rupavahini2024

I Am Freelancer

Featured on the national broadcaster to share practical insights on digital freelancing and how technology is reshaping the future of work for Sri Lankan youth.

WebinarOnline

Earning via Mobile App Development & Google AdMob

Shared practical knowledge on monetising mobile apps through Google AdMob with the OECSL community — part of an ongoing series of digital entrepreneurship webinars.

Start a Conversation

Have an AI challenge, a research collaboration in mind, or want to discuss a project? I read every message.

Good fits for a conversation

Privacy-preserving AI for your product or institution
Agentic RAG or on-device LLM architecture
NLP system design from research to deployment
Research collaboration or manuscript co-authorship
AI readiness assessment for your organisation

Find me on

MadusankaPremaratne

The Research behind the work

Core Expertise

Research Focus

Industry Application

Recognition