Jun Huang

Jun HUANG

Research Software Engineer · MSc CS (Georgia Tech) · PhD (EPFL)

I build prototypes end to end: web front ends and back ends, on-device ML, containerized cloud deployments and Unity/C# 3D applications. I work daily with agentic AI coding tools, and my research background brings experimental design and careful data analysis to the software I make.

Based in Zurich. Outside of code: hiking in the Alps, skiing and photography.

Software Projects

On-device ML Android apps that notify you when the sound level crosses a threshold or a TinyML model detects an alert sound, plus a companion app for real-time YAMNet audio classification.

AndroidJavaKotlinTensorFlow LiteYAMNet

3D horror-puzzle game: navigate trapped rooms, solve puzzles and escape a ghost. Built an AI chase agent, player-health system and UI. Team project.

UnityC#Game AI

Full-stack skin-cancer diagnostic web app: upload a lesion image, get a prediction from a CNN trained on ISIC. Full CRUD on patient and condition records via a HAPI FHIR R4 server; containerized and deployed on AWS EC2. Team of three.

PythonFlaskStreamlitTensorFlowFHIRDockerAWS

Smart task scheduler built in one day by a team of four at the IBM × GDG Zurich “Bobathon” hackathon (Apr 2026). Mood- and goal-based task generation, priority scheduling without overlaps, ICS calendar import and in-browser persistence.

JavaScriptHTMLCSSHackathon

Chrome extension that adds multi-select bulk deletion to the Gemini sidebar, kept in sync with the dynamically re-rendered page via a MutationObserver.

JavaScriptChrome ExtensionDOM

Background

Skills
Education
MSc Computer Science · Georgia Institute of Technology (GPA 3.9/4.0)
2023 – 2026
PhD Neuroscience · EPFL, Lausanne
2018 – 2022

Neuroscience Research

Research experience

Neural mechanisms underlying chronic pain

Neuronal Plasticity Group, University of Bern · Dec 2022 – Apr 2025

Miniscope, fiber photometry and patch clamp, focusing on the ACC–PAG circuitry; built custom fiber photometry setups and analyzed 20 TB of imaging data.

Neural mechanisms in reward-based learning

Laboratory of Behavioral Genetics, EPFL · June 2018 – Nov 2022

Individual differences in reward-based learning, focusing on dopamine dynamics. Open-sourced the Python/MATLAB data pipelines.

Late-associativity in auditory–amygdala circuits

CAS Key Laboratory of Brain Function & Diseases, USTC · Sept 2015 – Dec 2017

Temporal requirements for fear-conditioning association.

Psycho-physiology recording platform for rodents

Cell Electrophysiology Laboratory, Wannan Medical College · Nov 2010 – June 2013

Synchronized recording of central neural activity and peripheral functions.

Publications
High frequency optogenetic activation of inputs to the lateral amygdala forms distant association with foot-shock
Li, F., Jia, C.H., Huang, J., Bi, G.Q., and Lau, P.M., 2020. Molecular Brain, 13(1), 44.
Single morphine exposure distinctively enhances morphine-sensitivity of amygdala and nucleus accumbens in rats
Cheng, X.C., Zhang, S., Huang, J., Wang, W.W., Liu, W., and Wang, M.Y., 2015. Chinese Journal of Clinical Pharmacology and Therapeutics, 2015(09), 998–1003.
Simultaneous recordings of electrical activities from dual brain nuclei and multiple physiological functions in adult rats
Liu, W., Wang, W.W., Huang, J., and Wang, M.Y., 2013. Acta Academiae Medicinae Wannan, 32(5), 423–429.
My biggest inspiration is Rafa Nadal.