پرنیان سمسار
code, cosmos, caffeine, claviers.
About Experience Projects Research Contact

hello, world.

// parnian here.

CS student at the University of St. Thomas — I build ML models for astrophysics research and nerd out about creating software at night.

See my research Projects
Py
PT
Re
PS
import astropy as ap
model.fit(X_train)
SELECT * FROM stars
git commit -m "🚀"
scroll
/

about me

I was born and raised in Tehran, Iran. In 2023 I packed my life into two suitcases and moved to Minnesota — for the first time.

"I left home with so much love, I still think about it sometimes."

Baby Parnian on a keyboard

I'm currently a CS student at the University of St. Thomas, working as an undergraduate researcher in computational astrophysics, where I build ML and Deep Learning models, studying eclipsing binary stars using NASA TESS and Kepler data.

Parnian now

When I'm not in front of a screen, you'll find me at the piano, behind a telescope, or pitching ideas at business competitions.

Technologies I work with:

Python PyTorch Java React Native TypeScript
/* ↓ tap or hover */
Piano recital
UST Observatory
Pitching ACADS
/

experience

Undergraduate Researcher @ University of St. Thomas

JUL 2024 – PRESENT

Built automated Python pipelines for NASA TESS/Kepler light curve analysis — classical methods validated on 15+ stars (Mikulásek function + MCMC)

Automated manual methods with algorithmic pipeline achieving ~0.0017 day period error

Self-taught deep learning from scratch — designed and trained 1D-CNNs, RNNs, GRUs in PyTorch on 4,584 stars with 40M+ observations

Currently investigating preprocessing and catalog filtering strategies to improve model performance

PyTorchAstropyMCMC1D-CNNsGRUsPHOEBE

Observatory Crew Member @ University of St. Thomas

JUN 2025 – PRESENT

Setup, calibrate, and operate telescopes for public star parties and research

Capture and process astrophotography images of celestial objects

Assist visitors and provide tours of the UST Observatory

CS & Data Science Tutor @ University of St. Thomas

SEP 2024 – JAN 2025

Tutored peers in Data Science 120 and Computer Science 131

Helped students debug code and understand programming and problem-solving concepts

PythonCRJMP

Event Support Technician @ University of St. Thomas ITS

OCT 2023 – DEC 2024

Provided technical support for live campus events, including audio-visual setup and system troubleshooting

Resolved hardware and software issues in time-sensitive environments to ensure smooth event execution

Audio mixersHDMI/SIDPTZ cameraLive streaming setups
/

projects

📁

Smart Property Estimator

ML model for housing prices using Tehran district data. Scikit-learn regression.

Python · Scikit-learn · ML

📁

Web Tracking Project

Website with a custom Pillow-generated CAPTCHA, tracking visitor interactions via cookies, plus an admin event-log panel.

Python · HTML/CSS · Security

📁

Price Comparison Software

Desktop app comparing product prices and quality across retailers, with search and recommendations.

Python · Tkinter

/

research

I study eclipsing binary stars — two stars orbiting each other, periodically dimming one another's light. From their orbital period you can derive mass, and from mass, almost everything about the star.

My work: estimate that period automatically, at scale, across thousands of stars from NASA's TESS and Kepler missions. I started with classical methods (MCMC, Mikulasek) — precise but limited. Then I self-taught deep learning from scratch to scale up.

I tried 1D CNNs, GRUs, feature engineering, and preprocessing pipelines for ML training. Results are ongoing — but I've gone from manual, one by one analysis to an automated pipeline scaled to 4,584 binaries.

4k+
Stars analyzed
2yr+
Research tenure
2×
Best poster awards
PyTorch1D-CNNs GRUs / RNNsMCMC AstropyLightkurve PHOEBEELISA
Normalized TESS light curve of V454 Aurigae showing periodic eclipse dipsLight curve · V454 Aur · TESS
/

recognition

🏆

Best Research Poster — 1st Place

MINK-WIC · Kansas City · Oct 2025

First place at the Missouri, Iowa, Nebraska, Kansas Women in Computing Conference for undergraduate research in computational astrophysics and ML.

🥇

Best Research Poster — 1st Place

MNAAPT · Minnesota · May 2025

First place at the Minnesota section of the American Association of Physics Teachers for presenting my binary star research.

💡

Business Plan — Special Commendation

University of St. Thomas · Feb 2025

Presented ACADS — an AI-driven academic guidance platform. Recognized for innovation and business potential.

✦

Omicron Delta Kappa

National Leadership Honor Society · Nov 2025

Inducted for scholarship, leadership, and service across campus and community.

Say Hello
GitHub ↗ LinkedIn ↗ parniansemsar81@gmail.com