Arnau Andrews | Data portfolio

Arnau Andrews

About

I’m Arnau Andrews, a data analyst and data engineer with over 3 years of hands-on experience turning raw data into real business impact. With a background in International Business and a Master’s in Data Science, I specialize in building scalable ETL pipelines, analyzing complex datasets using Python and SQL, and delivering actionable insights through clear, compelling visualizations in Power BI and Tableau. I’ve worked across diverse industries — from public sector analytics to churn prediction models in e-commerce — and thrive at the intersection of data, business, and strategy. Fluent in English, Spanish, Catalan, and Italian, I’m always ready to bring data-driven thinking to international teams. Take a look at my portfolio to see the projects I’ve led, and feel free to reach out

Skills

Projects

Predictive Analytics: Investigation of customer buying patterns

Valuable insights were gained regarding customer behavior, including the relationship between items purchased and amount spent, online vs. in-store purchasing preferences, regional spending variations, and age differences. The analysis uncovered actionable insights to optimize sales strategies.

Time Series Analysis: Smart Home Power Consumption Analysis

Development and deployment of a comprehensive time series dashboard that enables Smart Home owners to analyze and visualize their power consumption data over time, facilitating informed decision-making and empowering them to adopt energy-saving habits with tailored insights and recommendations

Predictive location analytics: "Optimizing Indoor Localization: Evaluating Machine Learning Models Using a WLAN Fingerprinting Database"

This project offers a large database and methodology for indoor localization using WLAN fingerprinting. It improves predictions iteratively, with the Extra Trees Regressor excelling in longitude prediction. Overall, this project's applications include indoor navigation and location-based services.

Fraud Detection Analytics: Credit One Default Rate Analysis

Credit One Default Rate Analysis project aims to develop a data-driven solution to address increasing customer default rates. The analysis explores predictive models, feature engineering, and demographic variable analysis to understand creditworthiness, improve the loan approval process, and propose

Reccomendation Systems: Market Basket Analysis Tech Store

This report explores market basket analysis, uncovering valuable insights from customer purchase patterns in a tech store. The analysis uses association rules, generated using the R programming language and the arules package

Cluster Analysis: Identifying Patterns and Opportunities for Economic Growth for small businesses accros US States

This project utilizes the k-means algorithm to analyze small businesses in the United States and identify patterns and opportunities for economic growth. By clustering the businesses based on their economic data, we aim to uncover insights that can drive strategic decision-making.

Factor and Linear Discriminant Analysis: Wine Classification

This project explores the application of Factor Analysis and Linear Discriminant Analysis for wine classification. The findings reveal underlying characteristics related to citric acid and density, leading to improved accuracy and enhanced sensitivity for the minority class. These techniques offer v

HuffPost News Classification: Multiclass Classification with Naive-Bayes Algorithm and Web Scraping

This project utilizes a Naive-Bayes algorithm to classify news articles from HuffPost. With a dataset of 200,000 headlines, the objective is to identify categories such as Business, Parenting, Sports, Travel, and Wellness. Web scraping is used to gather extra headlines for evaluation purposes.

Business Analytics: Bike Store Sales Management Dashboard

This project uses Power BI to create detailed reports for marketing and sales teams. Its objectives include YTD reports, margin analysis, cost assessment, sales vs. LY comparison, and client/product categorizations. Power BI features like DAX functions, conditionals, switches, variables, and more.

Predictive Churn Analysis: Treatwell

The Treatwell Churn Prediction Project successfully implemented predictive models leveraging Lift Index and SHAP values to enhance venue churn prediction. By improving model performance metrics and identifying influential variables, Treatwell gained actionable insights to reverse churn trends.

Experience

01/2024 - Present

Data Analytics Consultant at SDG Group

• Oversee the operation of 1000+ tables, 200+ pipelines, and over 400 dataflows using Azure Data Factory, ensuring efficient ETL processes • Ensure efficient data movement, transformation, and integration processes through Azure Data Factory while leveraging GitHub for effective code management and collaboration • Proficient in creating dashboards, metrics, and models using PowerBI and DAX, with SQL for querying tables and database management

01/2023 - 01/2024

Data Scientist at Treatwell

• Processed data comprising 2M records with 50 + variables using Python. Ensured data quality through comprehensive cleaning, wrangling, and consistency checks. • Utilized statistical analysis techniques to identify critical factors influencing salon cancellations. • Employed both Shap Values and LIFT analysis for model interpretation and assessing effectiveness. • Leveraged both supervised and unsupervised machine learning techniques, including boosting and bagging to develop a robust Churn Prediction Model which captured a 300% increase in churn prediction accuracy.

02/2022 - 07/2022

Marketing Insights Analyst at Damm

• Compiled and meticulously analyzed EMEA beer market data from GlobalData, consistently delivering monthly reports with actionable insights. • Collaborated effectively with cross-functional teams to translate insights into targeted marketing strategies.

02/2021 - 07/2021

Business Data Analyst at STH

• Compiled and organized data into a database of 25.000 potential clients, improving lead generation efficiency by 30% • Conducted research on the kitchen sink industry, including market trends, customer demographics and factors influencing product demand using SQL in the SAF division.

09 / 2018 - 09-2020

Market Researcher Analyst

• Conducted market research and competitor analysis in the lightning protection industry, expanding the international customer database by 86% through targeted research efforts. • Utilized SQL and Excel to analyze pricing strategies, resulting in a 11.2% improvement in profit margins and effective pricing optimization.

Education

2022-2023

Master's Degree in Big Data and Data Science - University of Barcelona

Highly project-based program focused on big data tools and techniques, including machine learning, statistics, and advanced prediction methods. Grade: 8.9

2022-2023

Data Analytics & Machine Learning - Ubiqum Code Academy

Completed an 8-month on-site 800-hour project-based bootcamp at Ubiqum Code Academy, specializing in data analytics and data science.

2018 - 2022

International Business & Marketing - ESCI-UPF

• A degree which leads and equips students with the knowledge, expertise, techniques, skills and conduct that are the keys to success in international business and marketing.

2021

Exchange Semester at the University of Surrey, UK