I’m a Data & Business Analyst with 4+ years of experience owning end-to-end analytics, from raw data and transformations to insights and dashboards that support day-to-day decision making. I primarily work with SQL and Excel (including Power Query), have strong experience in Python (especially Pandas), and BI tools like Power BI and Tableau. I also bring 10+ years of hospitality operations experience, including leading multicultural teams of up to 40 people, which shapes how I partner with stakeholders today.
This portfolio showcases projects that I have been working on outside of my core role.
This project uses SQL-based data modeling and behavioral analysis to help Product, Growth, and Marketing teams understand how user engagement translates into purchases and where users drop off before converting. By transforming raw clickstream events into session- and user-level models and analysing cart behavior and checkout activity, the analysis supports decisions around checkout optimization, conversion strategy, and re-engagement tactics. The results translate directly into actionable recommendations such as checkout experiments, abandonment reminders, and more focused targeting of low-frequency buyers.
This project delivers an interactive Power BI dashboard that analyses restaurant sales data to support Restaurant Operations and Marketing teams in decisions around menu optimisation, promotions, and opening hours. Using Power Query for data cleaning and modeling and DAX for revenue and performance measures, the dashboard explores category- and item-level sales, meal periods, and weekday/hourly patterns. The analysis supports practical actions such as targeted lunch and happy-hour promotions, menu rationalisation, and adjusting opening and closing times to better align labour costs with demand.
This project is an exploratory Tableau dashboard built for the Maven Rail Challenge, where I acted as a BI developer supporting National Rail operations and commercial teams. Using data preparation, feature grouping, and interactive filters, the dashboard analyses route popularity, peak travel times, ticket-type revenue, and on-time performance. The results directly translate into recommendations: targeted operational fixes on high-impact routes, helping reduce refund requests and recover revenue from otherwise profitable connections.
This project uses Python-based data cleaning, exploration, and visualization to analyse historical Nobel Prize data and uncover patterns across gender, country, category, and institutions. Working in Jupyter Notebook with Pandas and multiple visualization libraries, the analysis supports questions around representation and long-term trends. The project demonstrates how exploratory analysis and visual storytelling can surface structural imbalances and concentration effects.
This project uses SQL-based exploration to analyse a delivery app dataset and identify popular cuisines, restaurants, and peak ordering times to support restaurant managers and delivery operations in staffing and scheduling decisions. The analysis examines order volume, basket size, delivery times, and customer ratings. The analysis supports practical decisions around shift planning and demand coverage during peak hours, while demonstrating core SQL querying and modeling skills on imperfect, real-world data.
This project uses Python-based exploratory analysis and multivariable linear regression to examine which factors most influence housing prices in 1970s Boston and to build a predictive pricing model. Working in Jupyter Notebook, the analysis explores feature distributions, correlations, and relationships across property, location, and socioeconomic variables before fitting and evaluating a regression model with transformed target variables. The price prediction model shows that room count, accessibility, and employment proximity positively affect prices, while higher crime and poverty rates depress values. The analysis demonstrates how structured feature analysis and regression modeling can be used to estimate property values and quantify the drivers behind price variation.
This project uses Python-based exploratory analysis and statistical testing to examine historical mortality data from the Vienna General Hospital and assess whether handwashing reduced maternal deaths. By analyzing yearly and monthly birth and death records from two clinics and comparing outcomes before and after hand hygiene was introduced, the analysis supports decisions around process change validation and causal inference. The analysis demonstrates how data can be used to validate operational changes and build evidence for adopting procedures that improve outcomes.