Applied AI Research · Privacy-Preserving LLMs · Agentic RAG
Madusanka Premaratne — Applied AI Researcher & Founder
Building production-grade, privacy-preserving AI architectures, semantic generalization pipelines, and on-device agentic systems. Founder & CEO of Knivok.
What I do
View all services →AI Privacy & Compliance Consulting
Design and implement query-time privacy frameworks for LLM applications — semantic generalization, data minimization, and GDPR-compliant AI pipelines without sacrificing utility.
AI Knowledge & Retrieval Systems
Build multi-agent retrieval-augmented generation systems for enterprise use — from architecture design through production deployment, including on-device and edge-first configurations.
AI Search & Text Intelligence Consulting
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.
On-Device & Data-Sovereign AI Consulting
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.
AI Strategy Consulting
Define your AI research agenda, evaluate model capabilities, and translate academic findings into product strategy. Research collaborations, manuscript preparation, and AI readiness assessments.
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.
Selected research & work
View all research →EdgeTal
On-device agentic RAG system for privacy-preserving talent discovery — full retrieval and inference pipeline at the edge, zero candidate data transmitted to the cloud. Presented at EICON 2026 (ESOFT International Conference), Kandy — awarded Best Poster Presenter.
From Surveys to Signals
An attribution-stability audit of machine-learning driver rankings on correlated HR survey data — testing whether SHAP-based explanations can be trusted, and building a deployable two-item scoring instrument from what actually holds up under resampling.
Founder & CEO at Knivok · Applied AI Researcher · Best Poster Presenter at EICON 2026 · Accepted Author at AIRC 2026 · 10+ Years Enterprise Software & AI Systems
Privacy-First LLM Systems
Query-Time Semantic Generalization
SemanticGuard's core method: sensitive fields are generalized at the moment of the query — before anything reaches a cloud LLM — rather than redacted after the fact. Evaluated on 32,593 queries at 95.1% data protection with 93.4% utility preserved.
On-Device Agentic RAG
EdgeTal's full retrieval, reasoning, and generation pipeline runs on local hardware. Zero candidate or user data is transmitted to any cloud inference endpoint — the architecture, not a policy, enforces this.
Data Minimization by Architecture
Collecting less, and generalizing what is collected, beats collecting everything and relying on access controls or post-hoc redaction to contain it.
Utility-Preserving, Not Just Private
Privacy techniques that gut model usefulness don't get adopted. The research question is the tradeoff itself — SemanticGuard measures both protection and utility simultaneously.
Edge-First Inference
Quantized models running on commodity hardware eliminate the round-trip to a third-party API entirely — no data residency ambiguity, no inference-time exposure to a cloud provider.
GDPR-Class Compliance by Design
When raw data never leaves the device, there's no cross-border transfer and no third-party sub-processor to disclose. Compliance follows from the architecture.
Why Privacy-First Architecture Matters
Data Leakage via Training
LLMs can memorize and resurface fragments of the data they were trained or fine-tuned on. Without an architectural safeguard, sensitive inputs carry real statistical extraction risk.
Regulatory Exposure
GDPR/CCPA-class regimes attach real liability to cross-border transfer and undisclosed retention. Architecture that avoids transmission avoids the exposure by design.
Trust & Adoption
Enterprises evaluating AI vendors increasingly ask where their data goes before what the model can do. An architecture that answers that cleanly is a competitive advantage.
Bridging the gap between academic research and production enterprise AI
I am Founder, CEO & Chief AI Research Officer at Knivok, leading research in query-time semantic generalization, privacy-preserving LLMs, and on-device agentic architectures.
My work translates theoretical privacy guarantees into deployable software for global enterprises, backed by first-class academic credentials from the University of Jaffna and London Metropolitan University.
Working on something? Let's talk.
Pick a call option that fits — a first exploratory conversation, a build to scope, or an AI readiness check.
Initial Consultation
A no-pressure first call to talk through your AI or product challenge and see if there’s a fit.
Development Services
Scope a build with the Knovik engineering team — LLM systems, agentic RAG, edge AI, and integrations.
AI Readiness Assessment
A structured assessment of where your organisation stands on AI adoption, and what to do next.