Sakuna Jayasundara

Hi, my name is

Sakuna.

I research & build AI stuff!

A passionate software developer and a researcher. I use advanced NLP and machine learning techniques to create intelligent systems that are intuitive, reliable, and user-friendly.

Resume

About Me

Hi there! I’m a final year Ph.D. candidate at the University of Auckland, where I’m researching human-centric cybersecurity. My focus is on developing AI techniques, particularly natural language processing (NLP) and GenAI, to translate high-level requirements into enforceable access control policies. I’m also dedicated to making cybersecurity systems more usable, ensuring people interact with them effectively and make fewer mistakes.

Before diving into academia, I worked as a software engineer at H2O.ai, building cutting-edge AI applications and developing robust testing frameworks. Prior to that, I was a backend engineer at ADL, crafting APIs with Java and Spring Boot.

I’m passionate about creating AI-driven solutions that bridge the gap between humans and technology. Outside of work, you’ll find me playing chess, enjoying cricket, or brainstorming my next big idea.

Here are a few technologies I've been working with recently:

PyTorch
PyTorch Geometric
Python
Tensorflow
Java
Huggingface

Experience

University of Auckland
H2O AI
Axiata Digital Labs
Paraqum Technologies

Graduate Teaching Assistant - University of Auckland

Oct 2022 - Present

While conducting research in my Ph.D., I am currently working as a graduate teaching assistant of the School of Computer Science, University of Auckland. I teach,

CS718 - Programming for Industry: Focusing on object oriented programming (OOP) as well as data structures and algorithms (DSA) in Java.
CS719 Programming with Web Technologies: Focusing on developing full-stack applications using JavaScript, Svelte, and SQL

Technologies I use to teach students are, Java, JavaScript, SQL

Software Engineer - H2O AI

June 2022 - Oct 2022

I worked as a software engineer at H2O.ai. My responsibilities involved,

Developing cutting edge AI applications using H2O Wave framework.
Developing end-to-end test automation pipelines for the applicatiions deployed in H2O Cloud.
Designing, developing, leading, guiding, and mentoring to implement a best-in-class test automation framework and continue to evolve the test automation roadmap.

Technologies I commonly used during my time at H2O are Python, PyTorch, Jenkins, PyTest, Plywright

Software Engineer - Axiata Digital Labs

June 2021 - June 2022

I worked as a software engineer at Axiata Digital Labs (ADL). My work at ADL involved,

Developing scalable backend APIs for the API Gateway in Celcom, Malaysia to provide necessary services for the customers all around the world.
Developing a PoC system to detect anomalies in already deployed services using open telemetry and machine learning
Developing a NLP-based application to detect public github repositories containing information related to the projects of the company.

Technologies I commonly used during my time at ADL are Java, Springboot, Python, PyTorch, Jenkins

Intern Electronics Engineer - Paraqum Technologies

June 2019 - Dec 2019

I started my career as an intern electronics engineer at Paraqum Technologies. My work at the company involved,

Developing a GTP (GPRS Tunneling Protocol) packet analyzing application with a Testing environment.
Developing a load balancing software to analyze network interfaces and manage data traffic.
Performance and functionality enhancement for the Active Directory (AD) Client software used in the company.

Technologies I commonly used during my time at Paraqum are C++, Verilog, Python

Education

2022 - 2025

Ph.D. in Computer Science

University of Auckland

Thesis: Access Control Policy Generation from High-level Requirement Specifications

Employed advanced deep learning and natural language processing techniques like large language models (LLMs) and retrieval augmented generation (RAG) to develop a novel relaible access control policy generation and verification pipline.
Introduced and developed a reliable conflict detection and resolution mechanism using graph neural networks (GNN).
Designed and empirically evaluated a usable access control policy authoring interface usability principles and explainable security concepts, the interface simplifies policy generation, reduces errors, and enhances user satisfaction.

2016 - 2021

Bachelor of Science (Hons.) in Electronics and Telecommunication Engineering

University of Moratuwa

GPA: 3.92 out of 4.2

Thesis: Object Detection, Tracking and Suspicious Activity Recognition for Maritime Surveillance using Thermal Vision

Won 2nd runner up in National ICT Awards (NBQSA) - Best university project
Developed an algorithm for Object Detection, Tracking and Suspicious Activity Recognition for Maritime Surveillance using Thermal Vision.
Developed a novel, state-of-the-art spatio-temporal activity detection framework utilizing key-point based detection architecture.
The developed system has the ability to do surveillance tasks with unmanned vessels and help navy personnel to detect suspicious activities in the sea.

Extracurricular Activities