kimball-methodology

Fortune-500-grade banking analytics platform: OLTP -> medallion lakehouse -> Kimball star schema -> semantic layer -> 9-tab executive dashboard + 5 ML models (churn, fraud, segmentation, forecasting). Production-ready, governed, fully tested.

A pipeline pulling from TMDB API that utilises Azure Data Lake Storage, Databricks/PySpark, JSON to Parquet transformation, Medallion Architecture and Kimball Modelling via dbt. Planned to be visualised in PowerBI for analytical work.

Master’s thesis by João Ramos (NOVA IMS), implementing a Kimball-based dimensional data warehouse and Power BI dashboards for Lisbon’s Na Minha Rua LX platform to enhance municipal decision support through a Microsoft Fabric Medallion architecture.

This was a guided project from the Analytics Engineering Bootcamp on Udemy

A hand-built, end-to-end ETL pipeline that extracts the classic Northwind dataset from a local PostgreSQL source, stages it in Snowflake, and transforms it into a Kimball-style star schema (with one deliberate snowflake branch) — orchestrated with Prefect and validated automatically before it's trusted.

My first data modelling project. Star Schema for Palladium Bank — built from 15 raw columns to a fully optimised, analytics-ready warehouse

This project is an automated end-to-end data pipeline designed to track and analyze stock market performance alongside media sentiment. By orchestrating the dataflow from the Polygon API through a dbt-transformed BigQuery warehouse, the system turns raw financial data into an actionable Star Schema.

ETL pipeline for Pakistan Stock Exchange (PSX) closing summary data. Star-schema warehouse with technical indicators (SMA, Volatility, Volume MA). Python, Pandas, SQLite, pytest.

This project takes the public GitHub dataset on BigQuery and models it into a Star Schema. It provides a single source of truth, enabling analysis of market share and open-source legal compliance.