Data Scientist · Computational Physics Researcher · Assistant Professor
From data to decisions. From equations to simulations.
Google Scholar · ORCID · Repositories
About / Fast Wave / Experience / Research / Toolkit / Credentials
A scientist's curiosity. A developer's discipline. A teacher's perspective.
I work at the intersection of data science, applied mathematics, and scientific computing. My experience connects banking applications, machine learning research, and numerical simulation.
I have developed forecasting and model-validation solutions for financial applications. During my master's research, I created Fast Wave, an open-source Python package for quantum harmonic oscillator wavefunctions.
As an Assistant Professor, I also teach mathematics and programming. Clear communication, reproducible work, and understanding the assumptions behind a model are central to how I work.
Simulating nature on a laptop still feels like keeping a small piece of the universe on my desk.
|
OPEN-SOURCE SCIENTIFIC SOFTWARE · MASTER'S RESEARCH
Quantum wavefunctions, computed efficiently. A Python package for calculating the position-space wavefunctions of quantum harmonic oscillator Fock states, with applications in photonic quantum computing. The project brings together numerical methods, accelerated computation with Numba and Cython, and arbitrary-precision calculations with mpmath.
|
Explore the research behind Fast Wave
The package was developed as part of my completed master's research. It reflects my interest in translating mathematical formulations into scientific software that can be inspected, tested, and reused.
| Area | Experience |
|---|---|
| Forecasting | Time-series modeling for bank budgeting and Expected Credit Loss estimation; comparison of statistical and deep learning approaches. |
| Model validation | Standardized validation pipelines, quantitative scoring, challenger models, performance evaluation, and governance criteria. |
| Risk analytics | Credit, operational, market, microcredit, climate, and environmental risk modeling and validation; financial measures including VaR and RAROC. |
| Machine learning & NLP | Neural networks, gradient boosting, clustering, and Transformer-based approaches to trajectory modeling. |
| Anomaly detection | Isolation Forest applied to event logs, with RabbitMQ and Elasticsearch in the supporting stack. |
My modeling workflow includes exploratory analysis, feature engineering, benchmarking, reproducibility, and evaluation. I am interested in both how well a model performs and the assumptions that support its use.
Methods and models I have worked with
- Time series: ARIMA, SARIMAX, Prophet, N-BEATS, N-HiTS, Autoformer, and FEDformer.
- Predictive and challenger models: MLP, XGBoost, LightGBM, and XGBSE.
- Clustering: KMeans, PAM, and CLARA.
- Anomaly detection: Isolation Forest.
- Natural language processing: Transformers and Hugging Face.
- Financial risk measures: Value at Risk (VaR) and RAROC.
My research has explored quantum simulation, deep learning, mobility data, and neuroevolution.
- Quantum wavefunctions: Efficient Computation of the Wave Function Using Hermite Coefficient Matrix in Python — WECIQ 2024.
- Smart farming: Research on deep neural networks for smart farming.
- Trajectory modeling: NLP-based trajectory modeling · TEACH research code.
- Trajectory classification: Deep learning for trajectory classification.
- Game AI: Neuroevolution research.
Full publication list and current citation metrics →
Core tools
| Focus | Tools and languages |
|---|---|
| Scientific computing | NumPy, Numba, Cython, mpmath, C/C++, MATLAB, Wolfram Mathematica |
| Data analysis & modeling | Python, R, SQL, Hugging Face |
| Distributed processing & data platforms | PySpark, Databricks |
| Databases & search | PostgreSQL, MongoDB, Elasticsearch |
| Systems & experimentation | RabbitMQ, Arduino, Assembly |
| Scientific writing | LaTeX |
- Databricks Fundamentals — Academy Accreditation
- Fundamentals of Deep Learning — NVIDIA
- Data Analysis in Databricks — DataCamp learning track
- Credit Risk Modeling in Python — DataCamp
Additional Databricks coursework
Language statistics reflect repository contents, including notebooks.
Good questions are a great place to start.
Interested in data science, scientific computing, or research collaboration?
Let's connect on LinkedIn → · Explore my portfolio