This repository contains tools, scripts, and documentation for the analysis and optimization of an Enterprise SSAS Tabular Model (Power BI Dataset).
The goal of this project is to programmaticly analyze the semantic layer of an existing SSAS model to identify performance bottlenecks, understand code dependencies (DAX/M), and propose state-of-the-art (SOTA) optimizations.
data/sample_metadata/: Contains dummy/anonymized metadata extracts (XMLA, JSON drops, Excel layer documentation).docs/: Optimization reports, architecture recommendations, and consultant ideas.scripts/: Python scripts used to parse, analyze, and extract insights from the XMLA/JSON metadata.notebooks/: Exploratory scripts and scratchpads for testing data parsing.
- Python: Used for parsing complex XMLA structure and Excel metadata (
pandas,json,xml). - Semantic Layer Optimization: Programmatic dependency mapping of layers and evaluating query performance.
- Reporting: Translating technical bottlenecks into actionable consultancy advice and migration strategies (e.g., Fabric Direct Lake).
- Install requirements:
pip install -r requirements.txt- Place the XMLA or metadata Excel files into
data/sample_metadata/(Please ensure all PII and sensitive data are removed). - Run the analysis scripts from the
scripts/folder:
python scripts/analyze_layers.pyImportant: The metadata files in this repository have been anonymized. No real PII, company figures, or sensitive internal business mechanics are included. For portfolio display purposes, these are sample structures.