Library

Everything you look things up in: the glossary, professional templates, printable cheat sheets, a curated catalog of external resources, a ready-made Power BI project and the practice datasets.

Study tools

Glossary

Professional templates

Requirements and open questions: What the requester needs, in numbers, with the questions still open and who answers them.
Assumptions log: Every assumption you made to keep moving, its impact if wrong, and who must confirm it.
KPI dictionary: One entry per business measure: definition, formula, grain, owner, source and caveats.
Validation and reconciliation: How you prove the numbers: source-to-report reconciliation, tests, tolerances and results.
UAT plan and log: Test cases the business runs before go-live, with expected and actual results and an outcome per case.
RLS / OLS test matrix: Every role and a representative user, what each must and must not see, and the result after each release.
Performance before/after report: Baseline, root cause per visual, fix, measured improvement and the budget it now meets.
Architecture decision record (ADR): Context, constraints, options, decision, consequences, risks and the trigger to revisit.
Pull request description: What changed, why, how it was tested, the risk and how to roll back.
Release notes and migration plan: What changes for users and when, the steps, the checks and the rollback.
Deployment checklist: Checks before and after promoting content between environments.
Runbook: Step-by-step diagnosis and recovery for a known failure, written for whoever is on call.
Incident report: The facts of an incident as it happens: impact, timeline, mitigation, status updates.
Blameless postmortem: Why it happened, why it wasn't caught, and the actions that stop it happening again.
Data contract: What a source team promises a consuming team: schema, meaning, freshness, quality and change notice.
Ownership matrix (RACI): Who is responsible, accountable, consulted and informed for each definition, model and process.
Data dictionary: Every table, column and measure users can see: meaning, type, examples, source and caveats.
Source-to-target mapping: Exactly how each target column is built from the source, with the rule, type and the test that proves it.
Semantic model design document: Purpose, facts and grain, dimensions, relationships, storage mode, security, measures and the performance budget.
Code review checklist: What a reviewer checks in a Power BI pull request: correctness, maintainability and impact.
Deprecation checklist: Retire a report or model safely: decide, announce, retire, then delete, reversibly until the end.
Capacity review: Monthly look at capacity usage, throttling and the top consumers, with actions and owners.
Workspace and content naming standard: Naming patterns for workspaces, models, reports, apps and security groups.
Certification checklist: The bar a semantic model must meet before it is certified: ownership, definitions, quality, security, usability.
Governance maturity assessment: Score ten governance areas with evidence, then pick three actions for the next quarter.
Center of Excellence charter: Mission, sponsor, responsibilities, scope and success measures for a BI Center of Excellence.

External resources

Power BI documentation: The authoritative entry point for Power BI Desktop, the Service, semantic models, reports, administration and sharing.
Power BI guidance: Microsoft's best-practice library: modeling, DAX, query folding, report design, performance, deployment and governance. Read it once you're past button-clicking.
End-to-end Power BI tutorial: Follows data from a raw source to a published report: a clean reference implementation before this course adds failures and ambiguity.
Report creation documentation: Visuals, formatting, navigation, filters, bookmarks, tooltips, AI visuals, mobile design and report samples.
Power Query documentation: Getting data, transformations and the Power Query editor.
Power Query M language: The functional language behind Power Query, for when generated steps aren't enough.
M function reference: Searchable reference for every M function: tables, lists, records, text, dates, web sources and errors.
Query folding fundamentals: How Power Query pushes work back to the source, and full vs partial vs no folding. Essential before DirectQuery and incremental refresh.
Query folding guidance for Power BI: Which transformations should fold, and when work should move upstream.
Query folding examples: No, partial and full folding demonstrated on a large table: good material for performance labs.
Query diagnostics: See what Power Query actually executes, the queries it sends to sources, and where authoring or refresh time goes.
Query folding indicators: Shows which steps fold in Power Query Online: useful for debugging transformation pipelines.
M error handling: try, otherwise, error records and robust error-handling patterns.
Star schema guidance: Facts, dimensions, grain and why star schemas make Power BI models faster and easier to use.
Model relationships: Cardinality, cross-filter direction, inactive relationships, referential integrity and how relationships really behave.
Create and manage relationships: The practical companion: setting cardinality and direction correctly in Desktop.
Many-to-many relationships: Bridge tables, facts at a higher grain such as targets, and other complex relationship scenarios.
Kimball dimensional modeling techniques: The professional reference for grain, transaction and snapshot facts, factless facts, surrogate keys, role-playing dimensions, bridges, late-arriving data, SCDs and conformed dimensions.
Get started querying with Transact-SQL: Free hands-on learning path: querying, joins and subqueries. A good prerequisite for the BI Developer and Engineer stages.
SQL Server documentation: The general SQL Server and T-SQL reference for source-side transformation and database work.
Execution plans: How the optimiser runs a query: essential when DirectQuery or extraction SQL is slow.
Compare and analyze execution plans: Before-and-after tuning comparisons and root-cause analysis of bad plans.
DAX documentation: The authoritative language reference.
DAX function reference: Syntax, parameters, return values and examples for every DAX function.
DAX syntax reference: Correct language structure, for when you move from copying measures to writing them.
CALCULATE function: Modified filter context, filter modifiers and context transition: the function everything else builds on.
Row context and filter context in DAX: A clear conceptual explanation of the two evaluation contexts.
Filter context in DAX: Why the same measure returns different results in different cells of a visual.
Filter context explained visually: A visual explanation for learners who find evaluation context abstract.
Understanding context transition: The bridge from intermediate to advanced DAX.
Context transition explained visually: A graphical explanation for when CALCULATE, iterators and calculated columns start to confuse.
DAX Patterns: Reusable solutions to business problems: time intelligence, budgets, inventory, segmentation, rankings and more. Free to read online.
SQLBI deep dive: Advanced DAX and semantic model material for after row context, filter context and context transition feel comfortable.
SQLBI filter context articles: Curated articles on KEEPFILTERS, ALL, SELECTEDVALUE and modern filtering practice.
The Definitive Guide to DAX, Third Edition: The book for genuine DAX mastery, with companion material.
DAX query view: Write EVALUATE queries, inspect the queries visuals send, test model results and debug unexpected numbers.
DAX Studio: The main external tool for querying semantic models and analysing DAX and model performance.
DAX Studio: Server Timings: Separates storage-engine from formula-engine work: central to the performance incident in Sprint 04.
DAX Studio documentation: Traces, query plans, VertiPaq Analyzer and model investigation.
Power BI reports overview: Visuals, interactivity, drillthrough, tooltips, bookmarks, navigation and accessibility.
Mobile layout view: Build dedicated phone layouts instead of assuming the desktop page works on a phone.
Mobile report best practices: Focus, sizing, spacing and storytelling for small screens.
Design accessible reports: Alt text, tab order, contrast, titles and reports that work with assistive technology.
Automatic page refresh: For operational and near-real-time reports.
Visual calculations: DAX calculations written on the visual itself: running sums, moving averages, comparisons.
PowerBI.tips theme gallery: Themes and layouts for inspiration. A community resource, not a design standard.
Share dashboards and reports: Sharing, workspaces, apps and how content reaches people, with the permissions each route gives.
Build permission for semantic models: Shared, reusable models and how report authors get controlled access.
Usage metrics: Find out whether a report is actually used before you optimise, change or retire it.
Configure scheduled refresh: Schedules, credentials, gateway mappings, refresh history and common failure causes.
On-premises data gateways: Standard, personal and virtual network gateways, and when each fits.
Gateway deployment guidance: Topology, machine sizing, placement and performance for enterprise gateways.
Troubleshoot gateways: Debug configuration, data source, semantic model and refresh failures, and collect logs.
Gateway implementation planning: Ownership, governance, availability and monitoring of gateways, for architects and administrators.
Incremental refresh overview: RangeStart and RangeEnd, partitions, and real-time hybrid policies.
Troubleshoot incremental refresh: Folding problems, the initial refresh, parameters and hybrid mode.
Row-level security: Static and dynamic RLS, roles, USERPRINCIPALNAME() and testing. RLS restricts Viewers, not workspace Admins, Members or Contributors.
Object-level security: Hide tables and columns from users entirely, rather than filtering rows.
Apply sensitivity labels: Microsoft Purview labels on reports, semantic models, dataflows and exported files.
Tenant-level security planning: Identity, groups, external users and administrative security decisions for the whole organisation.
Performance Analyzer: The first stop for finding which visual is slow, and whether the time is in DAX or rendering.
Optimization guide for Power BI: Microsoft's consolidated performance advice across reports, DAX, the model and sources.
Tabular Editor: Best Practice Analyzer: Automated model checks for naming, relationships, metadata, formatting, DAX and maintainability. Works as a model linter in CI too.
Composite models: Combining Import, DirectQuery and other storage modes in one model.
Direct Lake overview: Direct Lake on OneLake and on SQL endpoint compared with Import and DirectQuery.
How Direct Lake works: Transcoding, framing, automatic updates and DirectQuery fallback.
Develop Direct Lake semantic models: Current implementation guidance, including the difference between Direct Lake on OneLake and on SQL endpoint.
Analyze Direct Lake query processing: Diagnose fallback to DirectQuery and Direct Lake performance.
Power BI developer mode: The hub for PBIP projects, external tools, APIs, source control and CI/CD.
PBIP semantic model folder: How a model's definition, DAX queries and TMDL files are laid out in source control.
TMDL overview: The human-readable Tabular Model Definition Language used for model source control and collaboration.
Pro Git: The free, authoritative Git book.
Branches in a nutshell: Isolated feature work, and why BI changes shouldn't be made on the production branch.
Basic branching and merging: A realistic branch, hotfix and merge workflow, including simple merge conflicts.
GitHub: merge conflicts: Resolving conflicts, such as two people changing the same TMDL measure.
GitHub: pull request merges: Merge commits, squash merges and rebase-and-merge trade-offs.
Deployment pipelines: Dev, Test and Prod lifecycle management and controlled promotion of content.
The deployment process: Item pairing, stage behaviour, deployment rules and what gets copied between stages.
Automate deployment pipelines: Release automation through the API.
ALM Toolkit in an hour: Microsoft's hands-on workshop for comparing models, merging selectively and deploying metadata.
What is Microsoft Fabric?: The architecture at a glance: Power BI, Data Factory, Data Engineering, Data Warehouse, OneLake and the other workloads.
OneLake quickstart: Hands-on: getting data into OneLake, Delta tables and shortcuts.
Data Engineering in Fabric: Lakehouses, Spark jobs, notebooks and pipelines.
Lakehouse end-to-end tutorial: Builds a retail analytics solution with bronze, silver and gold medallion layers.
Fabric Data Warehouse: Warehouse use cases, T-SQL, star schemas on Delta storage, procedures and relational analytics.
Data Factory in Fabric: Ingestion, transformation and orchestration across Dataflow Gen2, pipelines, notebooks and SQL.
Dataflow Gen2: Power Query in the cloud for ingestion and transformation into Fabric.
Pipelines in Fabric: Orchestrate multi-step ingestion, transformation, SQL, notebook and control-flow processes.
Fabric Capacity Metrics app: Monitor capacity utilisation and find the items and operations that cost the most.
Optimize your capacity: Interactive and background workloads, throttling and finding resource-heavy operations.
Plan your capacity size: Capacity units, SKUs and estimating the capacity you need.
Fabric CI/CD documentation: The hub for Git integration, deployment pipelines, variable libraries and automation.
Fabric Git integration: Connect workspaces to Git and version-control workspace items.
Manage branches in Fabric: Isolated developer workspaces, branches and collaborative development.
Choose a CI/CD workflow: Git-based, build-environment, deployment-pipeline and API-driven release designs compared.
Power BI REST API: Automate content, administration, governance and embedding.
Semantic model (dataset) APIs: Refreshes, parameters, data source updates, DAX queries, refresh history and model operations.
Power BI Admin APIs: Tenant inventory, workspaces, models, users, activity events and labels for governance automation.
Fabric REST API: Automate items, workspaces, capacities, definitions, long-running operations and CI/CD.
Automate Git integration with APIs: Connect, commit and update workspaces from Git programmatically.
Semantic Link Labs: Microsoft's open-source Python library for semantic models, reports, capacities, Direct Lake, shortcuts, migrations and automation from Fabric notebooks.
Validate data with Great Expectations and semantic link: Microsoft's tutorial for automated expectations on semantic model data in Fabric.
Fabric adoption roadmap: The framework for growing from ad-hoc self-service BI to mature, governed analytics.
Governance (adoption roadmap): Balancing user empowerment with governance, security and organisational controls.
Content ownership and management: Choosing between business-led self-service, managed self-service and enterprise BI.
Center of Excellence: COE goals, responsibilities, team structures, mentoring and standards.
Adoption maturity levels: An organisational maturity model, beyond individual technical skill.
Power BI implementation planning: Strategy, tenant setup, workspaces, lifecycle, security, gateways, distribution and monitoring: one of the most useful sources for architects.
Tenant administration planning: Administration, domains, auditing, service monitoring and working with security and capacity teams.
System oversight: Governance, licensing, security, auditing, monitoring, cost and platform operations for BI leads and Fabric admins.
Track user activities: Retrieve and analyse activity events for governance, incident investigation and adoption analysis.
PL-300 study guide: Official exam objectives and change history; skills measured as of 20 April 2026.
Power BI Data Analyst Associate: Exam registration, renewal and official certification details for PL-300.
DP-600 study guide: The Fabric Analytics Engineer exam blueprint; the outline changes on 19 October 2026 (a minor change to "Query and analyze data").
Microsoft Fabric Community: Power BI forums: The large official community for Desktop, Service, Power Query, DAX, APIs, mobile and developer questions.
SQLBI: Specialist reference for DAX, semantic modeling, VertiPaq and performance.
DAX Guide: Function-by-function DAX reference with classifications such as iterators and context transition.
Tabular Editor documentation: Advanced model authoring, scripting, calculation groups and the Best Practice Analyzer.

Course datasets

DimDate: 90 continuous days, 1 Jan 2026 to 31 Mar 2026. Deliberately short: your first assignment is to replace it with a DAX date table.
DimProduct: 24 products in 4 categories. Includes a discontinued flag, unit cost and list price for margin calculations.
DimCustomer: 30 customers with segment, city, join date and loyalty tier. Some cities repeat across customers; join dates span 2022 to 2026 for cohort analysis.
DimRegion: 4 sales regions with the manager, target multiplier and country. The RegionKey is what FactSales, FactBudget and the RLS table join on.
DimEmployee: 18 employees in a parent-child hierarchy (ManagerKey points at another EmployeeKey). Built for PATH(), RLS by org position and hierarchy visuals.
FactSales: 100 order lines, Jan–Mar 2026. Multiple lines share an OrderID (for DISTINCTCOUNT), some lines are returns (negative quantity), and OrderDate ≠ ShipDate (role-playing dimension). Discount is a fraction of ListPrice.
FactBudget: 48 rows: monthly budget by Region × Category for Jan–Mar 2026. Different grain from FactSales (month, not day; category, not product). Built for budget vs actual and TREATAS.
FactInventory: 80 rows: weekly stock snapshot for 8 products × 10 weeks. A snapshot fact — summing QuantityOnHand across weeks is wrong. Built for semi-additive measures (LASTDATE, LASTNONBLANK).
RawOrdersExport (messy): 30 rows exported from a legacy system. Contains: header junk row, mixed date formats, currency symbols in numbers, trailing spaces, inconsistent casing, a fully null column, duplicate rows and "N/A" strings. Your Power Query playground.
SurveyWide (unpivot): 12 rows in a wide layout: one column per month. Built for Unpivot, and for the "why is this the wrong shape" conversation.
UserRegionMapping (RLS): 12 rows mapping a login to one or more regions. Some users have two rows (multi-region), one user has ALL. Built for dynamic RLS with USERPRINCIPALNAME().
ExchangeRates: 39 rows: weekly USD rates for EUR, GBP, CAD. Rates only exist on Mondays, so lookups need "last known rate" logic. Built for currency conversion patterns.
CustomerTargets (many-to-many): 20 rows: quarterly revenue targets by Segment and LoyaltyTier, not by customer. Neither column is unique, so it cannot join 1:many to DimCustomer. Built for bridge tables and many-to-many.
WebEvents (sessions): 100 rows of click events with SessionID, UserKey and timestamp. Built for funnel analysis, EARLIER/window functions, session duration, and "first purchase" cohort logic.

Starter Power BI project

Download the Northwind starter project (PBIP)

Library

Study tools

Glossary

Basics

Deployment

Architecture

Power Query

Fabric

Modeling

DAX

Visuals

Performance

Service

Governance

Security

Storage modes

Tools

Professional templates

External resources

Course datasets

Starter Power BI project