PORTFOLIO
PABLO CALVO - PERFORMANCE MARKETING & DATA ANALYSIS
Every piece of data has a story to tell – it is important that we listen.
Find more about my data journey down below.
MY DATA TOOLKIT
DATA STORYTELLING & VISUALISATION
With the rise of digital business and data-driven decision making, data storytelling has become a cornerstone skill for any organisation. Translating sophisticated data analyses into plain-English is paramount to provide stakeholders with valuable insights to propel the business forward. By simplifying and digesting billions of rows of complex data into an easy-to-understand chart, we empower our colleagues to make powerful and tailored decisions.
Data visualisation and data stoytelling gives anyone, regardless of level or skill set, the ability to understand and use data in their jobs every single day. Therefore, using data in an organisation should not be the goal – it should be part of the culture.
TABLEAU DASHBOARDING 
This Tableau dashboard shows a snapshot of an analysis on Pokémon Total Power. The data used is from Kaggle and includes information from every Pokémon from Generation 1 through 7. The raw data was processed in order to clean it, normalise it and add calculated fields used in the analysis, including the Total Power metric. The goal of the dashboard is to find a correlation between a Pokémon Total Power and other traits – Legendary condition, size or Type. Feel free to visit Tableau Public for a better experience.
GOOGLE DATA STUDIO DASHBOARDING
This dashboard showcases full-stack data processing – data is ingested via Python with web scraping, then normalised – cleaned, aggregated, additional calculated fields – and loaded into BigQuery (SQL Database). Lastly, via custom queries, data is pipelined into the dashboard for visualisation. The dashboard can be filtered by users and it’s completely dynamic – it’s updated automatically every 12 hours with new data.
MACHINE LEARNING #1: PLAYER SENTIMENT 🤖
In this GitHub repository you will find a machine learning project. The goal is to predict if a review of a game is negative or positive using an SVM model. You will find the training data and an additional, independent testing dataset to try the model.
There’s no need to run the script to get the model – you can download it directly from the ‘data’ folder, load it with your favourite Python IDE and test it with any text you want.
MONTE CARLO SIMULATION #1: MTG BOX GENERATOR 📦
A MTG Box Generator is used to calculate estimated value of boxes for different sets (Modern Horizons 2 in this example). It uses modelling for statistical analysis and helps users to gauge if the retail price of a box compares positively with the potential cards inside each booster pack.
You can find the latest version of the Box Generator and the data to run it in my GitHub repository.
CASE STUDY #1: RESERVED LIST IN EUROPE 📈
CASE STUDY #2: POKÉMON POWER CORRELATION
MY DATA ANALYTICS CERTIFICATES
