Selected Work

Personal Projects

Live

Personal Project · Node.js

Private Finance Dashboard

A local, privacy-first budgeting app that reads bank statements (CSV, Excel, PDF) straight from your computer. No cloud, no bank login, nothing ever uploaded. It parses statements from four different banks across two countries, categorizes spending, and tracks savings goals.

The part I'm most proud of: a monthly spending forecast built on least-squares linear regression per category, trained only on completed months so the current partial month doesn't bias the trend. It falls back to a plain average when the fit is too weak (r² < 0.3) or history is too short. A small but deliberate choice, to avoid presenting a shaky trend as a confident number.

Node.js Vanilla JS Linear Regression Time-series Forecasting 36 unit tests

View Code

The dashboard: running balance, balance history and monthly money flow — The dashboard. Running balance, daily balance history, and money flow by month.

Next month's forecast: predicted net, income and spending, broken down by category — Next month's forecast, per category. Net is derived from income minus spending, so the three figures always reconcile.

Spending tab: category breakdown and recurring monthly commitments — Spending by category for any month, plus the fixed commitments that repeat every month.

Goals tab: two savings goals with progress bars and feasibility against projected leftover — Savings goals, each checked against your projected monthly leftover rather than a flat average.

Running on fabricated statements, not real financial data. Click any image to enlarge.

In Progress

Personal Project · Python

arXiv Research Agent

An autonomous agent that monitors new arXiv papers in machine learning and NLP, summarizes them, and scores their relevance against a set of research interests: search, fetch, extract, summarize, and score, using real tool-calling rather than a single prompt-and-respond wrapper.

Python Claude API Tool Use arXiv API

Coming soon

University of St. Gallen

Academic Projects

Completed

Team of 4 · Completed June 2026

PikoGPT: A Compact GPT Built From Scratch

A 37M-parameter transformer built from scratch with a team of four: architecture, pre-training, and fine-tuning, with no pre-trained weights. I worked on the post-training stage, comparing two supervised fine-tuning strategies.

Hands-on experience in deep learning, training optimization, and model benchmarking.

Python Transformers Pre-training Supervised Fine-Tuning Benchmarking

Poster presented at the University of St. Gallen, June 2026. Click to enlarge.