Sajal Gupta
Ph.D. candidate in astrophysics @ CU Boulder
Sajal Gupta
About Me
Astrophysicist in training, studying how tilted accretion disks dance around spinning black holes.
I turn GRMHD simulations into physical models—via statistics, visualization, and machine-learning flavored inference.
I enjoy building tidy data pipelines and predictive models, then losing an argument to my own edge cases.
Off-duty: hiking and backpacking, learning trading the expensive way, and being a proudly mediocre programmer.
- Age 28
- Residence USA
- Address Boulder, Colorado
- Department Astrophysical & Planetary Sciences
- University CU Boulder
- e-mail sajal.gupta@colorado.edu
- Phone +1 720 742 6575
What I Do
Fun Facts
Working Hours/Day
10Projects
10+Research Papers
7+Resume
Education
Aug 2020 - May 2026
University of Colorado BoulderPh.D. in Astrophysics (Expected)
Thesis: Investigating the Dynamics and Observational Signatures of Misaligned Accretion Disks Around Black Holes. Advisor: Jason Dexter. GPA: 3.92
Aug 2015 - May 2020
IISER Kolkata, IndiaBS-MS Dual Degree in Physical Sciences
5-Year Integrated Program. First Class with Distinction (89%). Thesis: Solar wind interaction with a planetary off-center mini magnetosphere.
Experience
Aug 2020 - Present
CU Boulder & JILAGraduate Research Assistant
Leading computational research on relativistic plasma dynamics around supermassive black holes. Analyzed large 3D GRMHD simulation datasets (> 10 TB) to study alignment vs. precession in tilted black-hole accretion flows. Derived and utilized angular-momentum evolution to isolate the dominant physical drivers, and and integrated ML + signal processing techniques for scalable trend extraction and forecasting.
2025
Independent ProjectCustom LLM Development
Designed a privacy-preserving local LLM workflow that ingests a closed set of technical documents and returns citation-aware answers, summaries, and drafts. Built it to be scalable across corpora via structured ingestion, semantic retrieval, and reusable prompting templates.
Aug 2015 - May 2020
IISER KolkataM.S. Thesis Research
Ran 3D compressible MHD simulations (PLUTO-based module) of Sun–Mars interaction with an off-center weak dipole to test how localized magnetic structure shapes plasma flow, reconnection signatures, and magnetotail dynamics. Quantified boundary locations and escape outcomes (e.g., bow-shock standoff) and showed that mini-magnetospheres reduce atmospheric mass-loss relative to the no-dipole case.
2020 - Present
Academic & Side ML ProjectsApplied Modeling & Meaching Learning
Combined statistical inference on scientific datasets (e.g., χ²-style model fitting and parameter constraints) with applied ML coursework, including an end-to-end Kaggle project forecasting Black Friday sales using feature engineering and tuned models beyond baseline regression.
Programming / Data
Python
SQL
MATLAB
R
C++ / Fortran
ML / Scientific Stack
NumPy / SciPy / SymPy
pandas
scikit-learn
PyTorch / TensorFlow
Tools / Visualization
LaTeX
Matplotlib/seaborn/Tableau
Jupyter/VS code
Bash / Linux
Git
Plotly/VisIt/ParaView
Knowledges
- Data pipelines/RAG
- Feature engineering
- Model evaluation
- Time-series analysis
- Data visualization
- Reproducible research
- Version control (Git)
- HPC / SLURM
- Linux tooling
- Technical communication
DS/ML Courses
- Advanced Linear Algebra
- Advanced Mathematical Methods
- Probability & Statistical Learning
- Data Structures and Algorithms
- Applied Machine Learning
- Portfolio optimization
Academic Courses
- Advanced Statistical Methods
- High-Energy Astrophysics
- Fluid & Magnetohydrodynamics
- Radiatve/Dynamic Processes
- Atomic and Molecular Processes
- General Relativity
- Advanced Electrodynamics
- Advanced Quantum Mechanics
- Basics of Field Theory
