- π Graduate of BSc Computational Social Science, University of Amsterdam
- π¬ Focus: ML, NLP, data science, and applied AI, Business applications
- π§ I turn messy real-world data into models and models into working products
- Languages: Python, JavaScript, SQL
- ML / Data: pandas, NumPy, scikit-learn, statsmodels, XGBoost, Transformers (FinBERT), Ollama / RAG
- Web / Infra: Flask, FastAPI, Jinja, Tailwind, SQLite, Azure, Docker, Git
π― Application Tracker
Local-first job-application tracker I built during my own job search. Paste a posting URL and company / title / location are prefilled from the posting itself (ATS APIs β JSON-LD β heuristics); a plain Excel workbook stays the database; applications land as pins on a satellite globe. FastAPI + vanilla JS, 70 offline tests, CI.
Dual-sided ed-tech platform + a fully-local RAG tutor. Led the technical build.
Machine Learning β organised by technique (BSc coursework, UvA 2024β25):
- π£οΈ natural-language-processing β FinBERT sentiment, LDA topic modelling, VADER, TF-IDF, web scraping
- π supervised-machine-learning β regression, regularization, XGBoost / ensembles, a neural net from scratch
- β±οΈ time-series-forecasting β ARIMA, stationarity, ACF/PACF (COVID, Bitcoin)
- π§© unsupervised-and-causal-inference β PCA / t-SNE / UMAP, clustering, propensity-score matching
- πΌ LinkedIn: linkedin.com/in/djamie
- π§ Email: jamie.dongjae@gmail.com
π Amsterdam