Microsoft Fabric Essentials: No-Code to Pro-Code with Dataflows Gen2

#fabcon2026 #session #learning #microsoft #fabcon2026-session

Description

Microsoft Fabric Dataflows Gen2 empowers everyone to prepare data efficiently. In this session, learn no-code steps for quick wins, low-code techniques with Power Query for flexibility, and pro-code customization with M for full control. Discover practical strategies to simplify data prep and enable governed, reusable data at scale.

Key Takeaways

Complex multi-step orchestration (use Pipelines)
Custom code-first workflows (use Notebooks / Spark)
Copy-paste queries with no reuse model
One platform, multiple build levels, power and insightful UI. Choose the tier that fits your skill and use case
Power Queries with everything you need to get started and zero M language required
Configuring the output destination to a lakehouse table
Native lakehouse & warehouse support

My Notes

Action Items

Resources & Links

Slides

📥 Download Slides

Microsoft Fabric Essentials
No-Code to Pro-Code with Dataflows Gen2
Eric Overfield
CIO, Creospark, Microsoft MVP, Microsoft Regional Director
@ericoverfield
Eric Overfield
Chief Innovation Officer , Creospark
Eric Overfield
▪ Microsoft Regional Director
▪ Microsoft MVP, Microsoft 365 Apps & Services
▪ PnP Team Member
▪ Technology Community Organizer & Contributor
▪ Data and AI Innovator
ericoverfield.com
@ericoverfield
Enterprise advisory | Enterprise integration | Migration
Consultation
What we do
at Creospark.
Modern Work
Process
Transformation
Secure Cloud
We're Microsoft 365 implementation
experts passionate about
technology in pursuit of the
ideal digital employee experience.
AI readiness, Training, Adoption, & change management
Managed Services
Davis, California
Toronto, Canada
Our roadmap
~10 min
~30 min
~10 min
~10 min
Context & Tiers
Live Demos
Governance
Q&A
Dataflows Gen2
No-code, low-code,
and pro-code
Want to see this live?
Certification, lineage, and
git integration
Open discussion and
resources
Dataflows in a nutshell
ETL engine born in Power BI, evolved through Azure Data Factory, now native to Microsoft Fabric
Great at
Not designed for
✓
Connecting to your data - 150+ built-in connectors
✗
Real-time / streaming data ingestion
✓
Empowering citizen data engineers (no-code to pro-code)
✗
Complex multi-step orchestration (use Pipelines)
✓
Visual data transformations via Power Query (M language)
✗
Heavy compute transformations on massive datasets
✓
Deliver data directly into lakehouses & warehouses (v2)
✗
Custom code-first workflows (use Notebooks / Spark)
✓
Scheduled & incremental refresh for repeatable ETL
✗
Direct API serving or reverse ETL back to sources
Why Dataflows Gen2?
The Problem
The Fabric Answer
Fragmented data prep across tools and
teams
Ungoverned pipelines with no lineage
Unified data prep in one platform
(OneLake)
→
Built-in governance, lineage, and
endorsement
Skill-gap bottlenecks, ETL limited to devs
Multiple skill tiers: everyone can contribute
Copy-paste queries with no reuse model
Reusable, certified dataflows at scale
The capability spectrum
One platform, multiple build levels, power and insightful UI. Choose the tier that fits your skill and use case
NO-CODE
Business analysts, citizen data engineers
LOW-CODE
Power BI pros, experienced analysts
PRO-CODE
Data engineers, M language developers
Visual editor, drag-and-drop transformations, connect, and
publish
Power Query: custom columns, merge/append, parameters,
conditional logic, combine drag-and-drop with some M
Query
Custom M functions, dynamic sources, API patterns, error
handling, git integration
Git integration at all tiers
No-Code
Dataflow Gen2
Quick wins for everyone
No-code capabilities
Power Queries with everything you need to get started and zero M language required
Connect to Sources
Shape & Transform
400+ connectors: Data warehouses,
Lakehouses, SQL, SharePoint, REST
APIs, files
Visual editor: filter rows, remove
columns, change types, split/merge data
tables. Step through the transformation
process
Publish to OneLake
Monitor & Refresh
Output lands directly in your Lakehouse,
Warehouse, SQL DB, Azure SQL,
SharePoint, Data Lake Gen2
Governed and discoverable
Scheduled refresh, run history, and basic
lineage out of the box
DEMO
Building a Dataflow from scratch
Using only the visual editor. M code handled for us
WATCH FOR
Connecting to data sources and navigating schemas
Applying visual transformations: filter, rename, change type
Visual Editor
Configuring the output destination to a lakehouse table
Low-Code
Power Query for flexibility
Level up
Power Query: Power BI (Gen1) vs Fabric (Gen2)
Dataflow Gen1 (Power BI)
Dataflow Gen2 (Fabric)
— Power BI service only
Fabric workspace => unified platform
— Output to Power BI datasets
Cross-workspace access
— Limited data destinations
Native lakehouse & warehouse support
— No native lakehouse support
Direct Lake Mode, Fast Copy, and Staging
— Basic refresh scheduling
Advanced orchestration via pipelines
— No git integration (besides .pbix)
Full git integration & CI/CD
Key low-code transformations
Custom Columns
Create calculated columns using the Power Query formula bar. Date math, string ops, conditional logic
Merge & Append
Join tables on keys (inner, left, full) or stack datasets together. The backbone of data modeling
Conditional Logic
If/then/else columns, replace values, error handling. Business rules applied visually
Parameters
Dynamic values for connection strings, file paths, or filter criteria. Build once, reuse everywhere
DEMO
Power Query transformations
Parameters, merge, and reusable queries
WATCH FOR
Creating and using parameters for dynamic source selection
Merging two tables with a left outer join on a shared key
Referencing queries to build a reusable transformation chain
Power Query
Pro-Code
M Language for full control
Full code with CI/CD
When to drop into M
~20%
of scenarios need custom M
Custom Functions
Reusable logic across queries and dataflows.
Dynamic Logic
Parameterized sources, conditional paths, runtime-generated queries
but those are the high-value ones
API Patterns
Pagination, OAuth flows, nested JSON flattening. Real-world integrations
Error Handling
Try/otherwise blocks, graceful fallbacks, data quality guardrails
M Language Essentials
// Custom function: Clean & validate a text column
let expression
let
CleanText = (inputText as text) as text =>
let
trimmed = Text.Trim(inputText),
cleaned = Text.Clean(trimmed),
result = if Text.Length(cleaned) > 0
then cleaned
else "N/A"
in
result,
// Apply to a table column with
18 error handling
Source = Lakehouse.Contents(null),
Applied = Table.TransformColumns(Source,
{{"Name", each try CleanText(_) otherwise "ERROR"}})
function syntax
error handling
in
ATLANTA
MARCH 16 - 20, 2026
Applied
query folding
DEMO
Custom M functions
Dynamic source paths, parameterized API’s, error handling
WATCH FOR
Writing a custom M function from scratch in the Advanced Editor
let
GetPage = (n) =>
Json.Document(
Calling a REST API with pagination and dynamic URL construction
Web.Contents(
url,
[Query=
Wrapping calls in try/otherwise for graceful error handling
[page=n]]
)
in
GetPage
Making it real
Bringing IT together
Governance, reuse, and scale
Governed, reusable data at scale
Certified Dataflows
Git Integration
Data Lineage
Endorsement model lets you mark
trusted, production-ready dataflows
Version control, branching, pull
requests. Real CI/CD for data prep
End-to-end visibility:
source → transform → destination
OneLake Integration
Endorsement Model
Scheduled Refresh
Single storage layer. All outputs land in
one governed lake
Promoted → Certified → track trust
levels across your org
Automated pipelines with monitoring,
alerts, and retry logic
DEMO
Lineage and certified patterns
Governance in action
WATCH FOR
Navigating the lineage view to trace data from source to report
Endorsing a dataflow as Certified and viewing trust signals
Connecting a dataflow to a git repo and viewing change history
What is right for you?
What's your scenario?
Connect source, basic
transformations, publish?
Joins, custom columns,
parameters, or reuse?
Custom functions, APIs,
dynamic logic, error handling?
NO-CODE
LOW-CODE
PRO-CODE
Visual editor only
Power Query + formula bar
M language / Advanced Editor
Best for: analysts, quick data
preps, exploration
Best for: Power BI pros, data
modeling, reusable flows
Best for: data engineers, APIs,
complex logic
Remember: you can mix tiers in the same dataflow. Start no-code, drop into M only where needed
Key takeaways
One platform, many tiers of power
Power Query is your superpower in Fabric
Governance is built in, not bolted on
Dataflows Gen2 meets every skill level. No-code for quick wins, low-code for flexibility,
pro-code for full control.
The same PQ skills from Power BI translate directly, with dramatically more capable outputs and
destinations.
Certification, lineage, git integration, and OneLake make your data prep enterprise-ready from day
one.
Resources
Resources & Q&A
learn.microsoft.com/fabric/data-factory
learn.microsoft.com/powerquery-m
community.fabric.microsoft.com
github.com/microsoft/fabric-samples
Questions?
Thank you!
Microsoft Fabric Essentials
No-Code to Pro-Code with Dataflows Gen2
Eric Overfield
CIO, Creospark, Microsoft MVP, Microsoft Regional Director
@ericoverfield
Sound off.
The mic is all yours.
Influence the product roadmap.
Join the Fabric User Panel
Join the SQL User Panel
Share your feedback directly with our
Fabric product group and researchers.
Influence our SQL roadmap and ensure
it meets your real-life needs
https://aka.ms/JoinFabricUserPanel
https://aka.ms/JoinSQLUserPanel