Negar Hashemi

Postdoctoral fellow at University of Auckland

My research focuses on Artificial Intelligence for Software Engineering, software maintenance, testing, and software quality. I am particularly interested in building trustworthy AI-supported tools that help developers improve the reliability, maintainability, and performance of real-world software systems.

About

I am a Postdoctoral Fellow at the University of Auckland, working on AI-Driven Non-Functional Software Maintenance with Valerio Terragni and Kelly Blincoe.

My research focuses on Artificial Intelligence for Software Engineering, software maintenance, testing, and software quality. In my current work, I explore how AI, particularly Large Language Models and AI agents, can support non-functional maintenance tasks such as improving performance, maintainability, and reliability.

A key part of my research is understanding how AI-generated code changes can be made not only useful, but also correct, measurable, and trustworthy. I am especially interested in approaches that provide evidence that suggested software changes preserve functionality while improving important quality attributes.

During my PhD, I studied flaky tests in JavaScript projects, with a focus on identifying and mitigating order-dependent and environment-dependent test failures. This work shaped my broader interest in building reliable software systems and developing practical tools that support developers in real-world settings.

Alongside research, I value teaching, mentoring, and inclusive education. I have experience teaching software engineering and supporting students from diverse backgrounds, and I am always happy to connect with researchers and practitioners working in AI4SE, software maintenance, testing, and software reliability.

Software Engineering AI4SE Software Reliability Test Automation Automated Software Maintenance

Education

2026
Ph.D. Computer Engineering
Massey University · Palmerston North, New Zealand
Identifying and Mitigating Flaky Tests in JavaScript — Supervisor: Associate Professor Amjed Tahir
2013
M.S. Computer Engineering
Sharif University of Technology · Tehran, Iran
Extensions to Feedback Motion Planning — Supervisor: Prof. Mohammad Ghodsi

Teaching & Experience

2026
Professional Teaching Fellow
University of Auckland
Advanced Software Requirements Engineering (SOFTENG754) — lectures, practicals, and team-based project supervision.
2023–2025
Teaching · Teaching Assistant
Massey University
Software Engineering Design and Construction (159251) — lectures, tutorials, and marking.
2023–2026
Learning Interpreter (part-time)
Palmerston North Girls' High School
Personalised academic guidance and mentoring for diverse learners.
2016–2021
Research & Development Engineer
Megafa Company · Shiraz, Iran
Data management solutions (SQL Server, Crystal Reports), internal tools (Java, MySQL), access-control mobile apps and RFID systems.

Publications

SCP '26

JS-TOD: Detecting Order-Dependent Flaky Tests in Jest

N. Hashemi, A. Tahir, S. Rasheed, A. Shi, R. Blagojevic — Journal of Science of Computer Programming, Elsevier.

read paper →
ICST '26

A Systematic Evaluation of Environmental Flakiness in JavaScript Tests

N. Hashemi, A. Tahir, S. Rasheed, A. Shi, R. Blagojevic — IEEE Conference on Software Testing, Verification and Validation (ICST), Daejeon, South Korea.

ICST '25

Detecting and Evaluating Order-Dependent Flaky Tests in JavaScript

N. Hashemi, A. Tahir, S. Rasheed, A. Shi, R. Blagojevic — IEEE ICST, Napoli, Italy, 2025, pp. 13–24.

read paper →
ICST '25 · DS

Identifying and Mitigating Flaky Tests in JavaScript

N. Hashemi — IEEE ICST Doctoral Symposium, Napoli, Italy, 2025, pp. 779–781.

read paper →
JSS '23

Test flakiness' causes, detection, impact and responses: A multivocal review

A. Tahir, S. Rasheed, J. Dietrich, N. Hashemi, L. Zhang — Journal of Systems and Software, 206 (2023): 111837.

read paper →
ICSME '22

An Empirical Study of Flaky Tests in JavaScript

N. Hashemi, A. Tahir, S. Rasheed — IEEE ICSME, Limassol, Cyprus, 2022, pp. 24–34.

read paper →

Tools & Datasets

FlakyJS Dataset

Curated dataset of JavaScript projects exhibiting test flakiness, collected from open-source GitHub repositories.

view on Zenodo →

OD-Test Dataset

Dataset of order-dependent tests identified from open-source JavaScript projects.

view on Zenodo →

JS-TOD

Automated tool for revealing test order dependency in Jest through controlled reordering and rerunning.

view on GitHub →

js-env-sanitizer

Babel-based tool for mitigating environment-related flaky tests by automatically skipping and reporting them.

view on GitHub →

Env-Flaky Dataset

Dataset of environment-dependent flaky tests in JavaScript projects collected from open-source repositories.

view on GitHub →

Honors & Awards

Service & Leadership