This page is designed to showcase my complete & on-going projects in various categories, as well as my technical skills.
On-going Projects
Analysing years of Hungarian pharmaceutical shortage data to uncover market patterns, company‑specific trends, and ATC‑level insights — all compiled into an interactive Power BI dashboard for clear, data‑driven decision making. Currently designing a high-quality layout for the report dashboard.
Developed an end-to-end data pipeline to collect, clean, and structure job listings from HiringCafe into a structured Excel database, with ongoing development towards full scraping automation via Python and SQL integration.
Completed Projects
Developed an interactive Power BI dashboard using Azure Map visuals to geographically track hundreds of ongoing projects across Europe, with dynamic filtering by status, owner, and project type, fed by a daily-refreshed Excel data source.
- [Project Timeline Tracker (Gantt Chart)]
Built an interactive Power BI Gantt dashboard consolidating 100+ projects across multiple owners into a centrally tracked, auto-refreshing timeline with colour-coded milestones and dynamic filtering to replace scattered Excel-based manual reporting.
Built a dual-scenario Excel workforce capacity model for a regional audit hub, forecasting month-by-month audit throughput across a 24-person team through 2025 by accounting for staggered hiring, role-based training timelines, and vacation-adjusted realistic versus ideal capacity. Above you can find a dummy version of the model, as well as output of number of audits based on the models.
Conducted an exploratory analysis of a confidential European audit company list using Excel pivot tables and visualizations, uncovering geographic distribution patterns, company type breakdowns, priority category splits, and special audit requirements to support strategic resource planning.
Documented SQL query workflows and solutions across six mystery-themed SQL Noir cases, practicing database querying, table joins, and analytical thinking.
A hands-on learning log documenting my journey through the R for Data Science (2nd edition) workbook, covering data wrangling, transformation, and visualization using tidyverse, dplyr, and ggplot2 (snippet of graphs included above).
Practising and documenting solutions to real-world SQL interview questions from DataLemur, spanning pharmaceutical sales analysis, candidate filtering, assembly line tracking, and messaging analytics using PostgreSQL.
Completed a five-day SciLifeLab/NBIS workshop at Uppsala University covering the full RNA-seq analysis pipeline, from raw data quality control and read mapping through to differential gene expression and gene set enrichment analysis using R and Linux.
Coursework repository from the University of Colorado’s Clinical Data Science program, featuring projects spanning EHR-based hypertension phenotyping, ICU mortality risk prediction modelling, and NLP keyword extraction from clinical notes.
Documented a stakeholder email workflow using Excel functions (SUMPRODUCT, XLOOKUP, TEXTBEFORE/AFTER) to validate, extract, and structure contact data from merged strings for use in a customised Word mail merge campaign.
Cross-referenced and harmonized a dataset of ~200 Swedish ICD-8/9/10 disease codes against international WHO classifications for X-chromosome related autoimmune diseases at Linköping University’s Nestor Lab, identifying and documenting key coding discrepancies across classification systems using Excel.
Developed and implemented modular genomics and transcriptomics pipelines at Linköping University’s Nestor Lab, spanning LP-WGS alignment, RAP-seq RNA processing, and R-based Copy Number Variation analysis using Bash, Python, and R. Example above for expression profile from DNA CNV analysis.
Investigated the acute and chronic toxicological effects of alcohol-based and non-alcohol-based hand sanitisers on the freshwater indicator species Daphnia pulex using probit regression, Kaplan-Meier survival analysis, and heart rate endpoints in R, identifying alcohol evaporation as a key methodological confound.