Research
Fine-tuning LLMs for bug classification
Fine-tuned code transformers to sort GitHub bug reports into seven categories, benchmarked against classical ML on a hand-labelled corpus.
Open to full-time software & ML roles
Ardalan Askarian
Software engineer working on machine learning systems, and the ordinary software that has to hold them up.
M.Sc. in Applied Machine Learning, researching computer vision under Dr. Mark Eramian. B.Sc. Honours, Software Engineering option.
Turning research questions into working systems
Most of my work sits between a research question and the software that answers it: annotation platforms, data pipelines, and the interfaces researchers actually use. A good part of it is checking whether a thing works before claiming that it does.
I'm a Master's student specializing in Applied Machine Learning, researching computer vision and image processing under Dr. Mark Eramian. Before that, a B.Sc. Honours in Computer Science, Software Engineering option.
Where I've worked
Experience
May – Aug 2025
Research Assistant, NSERC USRA
Imaging & AI Lab
Built a Django annotation platform and ran a controlled study on whether SIFT-suggested bounding boxes actually help human annotators.
Oct 2024 – Sep 2025
Software Developer Intern
BEAP Lab
Led full-stack development of the BEAP Engine, a platform for ingesting, processing and analysing smartwatch sensor data.
Jan 2023 – Present
Teaching Assistant
Department of Computer Science
Six core computer science courses, including CMPT 332 Operating Systems.
Technical skills
Languages
PythonTypeScriptJavaScriptCJavaSQLC#R
Frameworks
ReactReact NativeNext.jsNode.jsDjangoFlaskPostgreSQLMongoDB
ML & computer vision
PyTorchTensorFlowOpenCVscikit-learnTransformersSIFTSegmentation
Tools
GitDockerUnix/LinuxPlaywrightUnity
Selected projects
Research first, then apps and games. Open any card for the detail.
Research · NSERC USRA
SIFT-assisted image annotation
An annotation platform built to answer one question honestly: does machine assistance make human annotators better, or just busier?
VR
Dreaming Machines
Biometric VR driven by a live pulse sensor.
Full-stack
BEAP Engine
Smartwatch data ingestion and analytics.
iOS
Weather App
Native Swift client for live forecasts.
Web
Dentistry Website
Booking flow and integrated maps.
Mobile
Sports Scheduling App
Led the front end for a team-management app.
Game
Darkness Defenders
Unity tower defence with layered enemy AI.
Let's talk.
I'm looking for full-time software and machine learning roles, and I answer every email.