rweisssieker-xp Power BI Modelling
Use this skill when source queries are known or a Power BI model must be designed or reviewed.
Workflow
- Define the business process and fact grain before visuals or DAX.
- Separate facts, dimensions, bridges, snapshots, and accumulating snapshots.
- Use conformed dimensions for date, organization, customer, product/material, account, geography, owner, status, scenario, currency, and source system where relevant.
- Define explicit measures with business meaning, numerator, denominator, date basis, exclusions, and reconciliation source.
- Design RLS/OLS and sensitivity labels early for customer, employee, supplier, finance, and regulated data.
- Specify refresh: import, DirectQuery, Direct Lake, incremental refresh, gateway, dataflows, deployment pipelines, and failure checks.
- Produce model documentation: tables, relationships, measures, refresh, security, validation, assumptions, and known risks.
References
- Load
references/model-patterns.mdfor model architecture and relationship rules. - Load
references/business-domain-catalog.mdfor common facts, dimensions, and measures by domain.
Assets
assets/semantic-model-spec-template.md: model specification template.assets/measure-catalog-template.csv: DAX backlog planning template.
Scripts
scripts/new_semantic_model_spec.py: Generate a semantic model starter spec.