Skip to content
View Om4143's full-sized avatar

Highlights

  • Pro

Block or report Om4143

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Om4143/README.md

From particle collisions to business insights

Hi, I'm Omveer 👋

Data Analyst & Data Scientist · PhD Physicist · turning complex data into decision-ready insights

📍 Darmstadt, Germany


About me

For 8+ years I analysed large, complex datasets in particle physics, first in my PhD, then as a postdoc at Goethe University Frankfurt working with GSI/FAIR. My job was to separate real signals from noise and to make sure the data could be trusted before anyone drew conclusions from it.

Now I bring that same rigour to business questions. I'm completing the Data Science Bootcamp at WBS CODING SCHOOL, where I turn raw data into clear recommendations with SQL, Python and Tableau.

  • 🔬 Analytical rigour - statistics and machine learning to find what's real in noisy data
  • ✅ Data quality - I built an automated QA framework that caught data errors before analysis
  • 🗣️ Communication - 20+ peer-reviewed publications explaining complex results clearly

Tech stack

Python Pandas NumPy SciPy scikit-learn MySQL BigQuery Tableau Google Cloud Git Docker R C++


Projects

📊 Data Analytics

A/B Test of Homepage Buttons · Python · SciPy Compared four button designs across ~100,000 homepage visits with a chi-square test and Bonferroni-corrected comparisons.

Result: white buttons roughly doubled the click-through rate of red ones (2.0–2.1% vs. 0.8–1.1%).

Discount Strategy Analysis · Python · Pandas Cleaned three raw sales tables (8,700+ duplicate rows, malformed prices) to test whether discounts pay off.

Result: sales stop rising above a 20% discount → recommended capping discounts and shifting them to high-end products.

Market-Entry Analysis · MySQL · Tableau Queried a 9-table marketplace database to decide whether an online tech retailer should partner with a Brazilian platform.

Result: ~93% of orders arrive on time and tech makes up ~14% of revenue → recommended signing the deal.

Smartwatch Activity Dashboard · BigQuery · Tableau Cleaned 960 days of my own Samsung Health data and built an interactive dashboard.

Result: my 10,000-step goal rate rose from under 30% to 73% in 2026.

⚙️ Data Engineering

Cloud ETL Pipeline · Python · Cloud SQL · Cloud Run · Cloud Scheduler Automated pipeline that scrapes city data from Wikipedia, pulls weather and flight data from two APIs, and loads it into MySQL on Google Cloud on a schedule.

📝 Read what broke and how I fixed it: Scraping, APIs and Cloud SQL: What My First Data Pipeline Taught Me

QA Data Framework · Python · C++ · CI/CD Framework that automatically checks large scientific datasets for errors, with quality checks built into the CI pipeline.


Find me elsewhere

LinkedIn Portfolio Email

Thanks for stopping by!

Pinned Loading

  1. samsung-health-analysis samsung-health-analysis Public

    I've worn a Samsung watch for a couple of years, so I exported the data and used it to practice.

  2. eniac-magist-analysis eniac-magist-analysis Public

    SQL analysis of Magist's marketplace data to evaluate a Brazil market-entry partnership for Eniac - tech-fit and delivery performance.

  3. eniac-ab-test-chi-square eniac-ab-test-chi-square Public

    A/B/n test analysis of homepage button variants for an online retailer. Chi-square test and Bonferroni post-hoc comparisons on click-through rate, using Python, pandas and SciPy.

    Jupyter Notebook

  4. eniac-discount-strategy eniac-discount-strategy Public

    A data analysis of Eniac's online store sales, built to answer one question: is discounting actually working, and how should it change?

    Jupyter Notebook