If you’ve worked with Azure Synapse, Azure Data Factory, Spark, or SQL-based ETL pipelines before, you’ll find many of the core concepts familiar. The key shift is learning how those concepts translate into Fabric’s integrated, SaaS-style environment.

Who Should Take This Exam?

According to the official Microsoft study guide, you’re the right candidate if your day-to-day work involves:

• Ingesting and transforming data using SQL, PySpark, or KQL
• Designing data architectures and orchestration workflows
• Securing and managing analytics solutions
• Monitoring, troubleshooting, and optimizing data pipelines

This exam is particularly valuable if you’re transitioning from the retired DP-203 (Azure Data Engineer Associate) certification. The concepts are largely the same; ETL/ELT, data transformation, data warehouse design, but DP-700 tests them through the lens of Fabric’s unified toolset.

Note:
DP-700 vs DP-600: If you’re more focused on Power BI, DAX, and analytics engineering, look at DP-600 (Fabric Analytics Engineer). DP-700 is strictly the data engineering path; SQL, PySpark, KQL, pipelines, and lakehouses.

Exam Format and Key Details

One important note: the exam is open book. You can access Microsoft Learn docs during the test. This doesn’t mean you can wing it. Scenario-based questions require genuine understanding, not just Ctrl+F searching, but it does mean you should practice navigating the docs quickly, especially KQL/PySpark syntax pages.

What Does the Exam Actually Test?

The exam is divided into three equally weighted domains, each comprising roughly 30–35% of the total score:

1. Implement and Manage an Analytics Solution (30–35%)

This domain covers the foundational setup and governance layer of Fabric. Expect questions on:

• Workspace configuration: Spark settings, domain settings, OneLake settings
• Lifecycle management: Git integration, deployment pipelines, database projects
• Security and governance: Workspace-level and item-level access controls, row/column/object-level security, dynamic data masking, sensitivity labels
• Orchestration: When to use Dataflow Gen2 vs. pipelines vs. notebooks; event-based triggers and scheduling

2. Ingest and Transform Data (30–35%)

The core data engineering domain. You need solid understanding of both batch and streaming patterns:

• Loading patterns: Full loads vs. incremental loads, dimensional model preparation, streaming ingestion
• Batch ingestion: Pipelines, shortcuts, mirroring, continuous integration from OneLake
• Transformations: Power Query (M), PySpark, T-SQL, KQL — knowing which tool to use in which scenario is key
• Data quality: Handling duplicates, missing values, late-arriving data, deduplication, aggregations
• Streaming: Eventstreams, Spark Structured Streaming, KQL windowing functions, Eventhouse vs. Lakehouse decisions

3. Monitor and Optimize an Analytics Solution (30–35%)

This domain tests operational maturity — what you do after the pipeline is built:

• Monitoring: Data ingestion, transformation, semantic model refresh, setting up alerts
• Troubleshooting: Resolving errors in pipelines, dataflows, notebooks, Eventhouse, Eventstreams, T-SQL, and shortcuts
• Optimization: Lakehouse table optimization, pipeline tuning, data warehouse query performance, Spark performance, KQL optimization

Best Study Resources for DP-700

1. Microsoft Learn Official Learning Path (Free)

This is your single most important resource. Microsoft’s official learning path maps directly to the exam objectives and includes both reading modules and hands-on labs inside a real Fabric trial environment.

Start here: DP-700 Exam Page on Microsoft Learn

Key modules to complete:

• Get started with Microsoft Fabric
• Implement a Lakehouse in Microsoft Fabric
• Implement real-time intelligence with Microsoft Fabric
• Implement a data warehouse in Microsoft Fabric
• Monitor and optimize data solutions in Microsoft Fabric

Pro tip: Since the exam is open book, actively practice navigating from the Learn homepage to Fabric docs, PySpark syntax references, and KQL documentation. Speed matters during the exam.

2. Official Microsoft Study Guide (Free)

Download the DP-700 Study Guide and use it as your checklist. Every bullet point in the “Skills Measured” section is a potential exam topic. Check each one off as you study it — this ensures you don’t have blind spots.

3. Free Practice Assessment (Free)

Microsoft offers a free official practice assessment on Microsoft Learn. It won’t give you the actual exam questions, but it accurately reflects the style, format, and difficulty. Take it early in your prep to identify weak areas, then again at the end to gauge readiness. This is an underused resource — don’t skip it.

4. YouTube Free Full Course

A solid free video course can help if you prefer watching over reading, this free YouTube course by Adam Marczak is a great option that cover the full exam syllabus, including Fabric architecture, lakehouse concepts, pipelines, notebooks, and real-time analytics. Watch the full video once, then revisit sections on topics where you feel less confident.

5. Udemy Course Paid But Worth It

For structured, in-depth coverage, Phillip Burton’s Microsoft DP-700 prep: Fabric Data Engineer Associate on Udemy is the top-rated option. It covers PySpark, SQL, KQL, and Fabric from scratch with no prior language experience required. Wait for a Udemy sale and you can get it for under $20..

6. Microsoft Applied Skills Labs

These are end-to-end scenario labs from Microsoft that simulate real Fabric workflows. Two are especially relevant:

• APL-3008: Implement a Real-Time Intelligence solution with Microsoft Fabric
• APL-3010: Implement a data warehouse in Microsoft Fabric

Completing these builds the hands-on confidence you need for simulation-style exam questions.

How to Get a Free (or Discounted) DP-700 Voucher

You can often secure a free or 50% discounted voucher by completing a Microsoft Cloud Skills Challenge (like “30 Days to Learn It”) or participating in the Fabric Community “Get Certified” program here. Additionally, check major events like Microsoft Ignite for limited-time free offers, or use the Enterprise Skills Initiative (ESI) portal if you work for a Microsoft Partner to see if your company covers the full cost.

Key Topics to Master (With Practical Focus)

Fabric Architecture: Know What Lives Where

You must clearly understand the purpose and appropriate use case for each Fabric item:

• OneLake — the single unified storage layer; think of it as Azure Data Lake built into Fabric
• Lakehouse — Delta Lake-based storage for open-format files; use for big data, ML, and flexible analytical workloads
• Warehouse — T-SQL-based, fully managed data warehouse; use for structured, enterprise-scale BI workloads
• Eventhouse — time-series and streaming analytics using KQL; replaces the old ADX in Fabric
• Eventstream — real-time data ingestion pipeline; routes streaming data to Eventhouse, Lakehouse, or other destinations

Lakehouse vs. Warehouse: The Most Common Exam Trap

Many candidates stumble on scenario questions that ask them to choose between a Lakehouse and a Warehouse. The general rule:

• Use a Lakehouse when you need open format storage, machine learning access, or heterogeneous data (files + tables)
• Use a Warehouse when the workload is primarily SQL-based, requires strong ACID transactions, or serves BI tools via a traditional SQL endpoint

Also understand the medallion architecture (Bronze → Silver → Gold) and when to apply it inside a Lakehouse.

Choosing the Right Transformation Tool

A frequent question type presents a scenario and asks which tool to use for transformation. Here’s the mental model:

• Power Query (M) / Dataflow Gen2 — best for low-code, analyst-friendly transformations; no coding required
• PySpark / Notebooks — best for large-scale transformations, complex logic, ML pipelines
• T-SQL — best for Warehouse-based transformations; familiar to SQL Server developers
• KQL — best for time-series and streaming data in Eventhouse

Incremental Loads and Slowly Changing Dimensions

The exam tests dimensional modeling knowledge more than most candidates expect. Make sure you understand:

• Full load vs. incremental load patterns
• Slowly Changing Dimensions (SCD) Types 1, 2, and 3
• Star schema design: facts, dimensions, surrogate keys
• Handling late-arriving and duplicate data

Real-Time Intelligenc, Don’t Skip This

Real-time analytics is a significant portion of the exam and is often under-studied. Focus on:

• KQL fundamentals: queries, window functions, aggregations
• Eventhouse architecture and when to use it
• Eventstream routing to different destinations
• Spark Structured Streaming vs. KQL-based streaming, know the trade-offs
• Accelerated shortcuts vs. non-accelerated shortcuts in Real-Time Intelligence

Security and Governance

Security questions are common and straightforward if you study them. Know the difference between:

• Workspace-level vs. item-level access controls
• Row-level security (RLS), column-level security (CLS), object-level security (OLS)
• Dynamic data masking — how and when to apply it
• Sensitivity labels and item endorsement
• OneLake security configuration

Practical 3-Week Study Plan

This plan assumes you have some prior experience with data engineering or Azure. Adjust the timeline based on your background.

Week 1 : Build Foundations

• Read through the DP-700 Study Guide and create a topic checklist
• Watch a full YouTube course covering the exam syllabus
• Sign up for a Microsoft Fabric trial and explore the interface
• Take the Microsoft Learn free practice assessment to identify your weak areas

Week 2 : Go Deep with Microsoft Learn

• Complete the official Microsoft Learn modules (don’t skip the hands-on labs)
• Build a simple Lakehouse → pipeline → Warehouse workflow yourself
• Practice writing basic KQL queries and PySpark transformations
• Work through the APL-3008 and APL-3010 Applied Skills labs

Week 3 : Review, Practice, and Exam Prep

• Revisit all weak areas identified in Week 1
• Take the Microsoft practice assessment again; aim for 80%+ before booking the exam
• Practice navigating Microsoft Learn docs quickly (for the open-book exam)
• Save KQL/PySpark/T-SQL syntax pages as bookmarks for exam day
• If you previously had DP-203, also take that practice test — there’s significant KQL and pipeline overlap

Exam Day Tips

• Run the Pearson VUE hardware check a week before — don’t discover your webcam or network has issues on exam day
• Bookmark key Microsoft Learn pages before the exam: PySpark syntax, KQL reference, T-SQL docs, Fabric pipeline documentation
• Read every scenario question carefully — the “best” answer often hinges on one specific constraint mentioned in the scenario (cost, latency, existing tools, team skillset)
• Don’t over-index on code syntax — the exam is more about choosing the right tool and architecture than writing perfect code
• Flag and move on — if a question stumps you, flag it and come back; don’t burn 10 minutes on one question
• DP-700 does NOT test Power BI or DAX — don’t waste study time on report building or visual design

Is DP-700 certification worth it in 2026?

Yes, and the timing is good. Microsoft Fabric is seeing rapid enterprise adoption, with a significant portion of Fortune 500 companies already on the platform. As organizations migrate from fragmented Azure data services (Synapse, ADF, ADX) to the unified Fabric platform, demand for certified Fabric Data Engineers is growing across finance, healthcare, e-commerce, and logistics.

The certification is also low-risk to maintain: renewal is free via a short online assessment on Microsoft Learn, keeping your credential current as Fabric evolves (which it does — monthly).

Final Thoughts

DP-700 is a well-designed exam that rewards genuine understanding over rote memorization. The open-book format means you can look up syntax, but you still need to know which tool to reach for in each situation, and why.

The most effective preparation combines three things: the official Microsoft Learn modules for structured coverage, hands-on labs inside a real Fabric trial for muscle memory, and the free practice assessment to measure your readiness honestly. You don’t need expensive courses to pass, but you do need to actually build things in Fabric, not just read about them.

Good luck and once you pass, remember to bookmark the renewal reminder. The free annual renewal keeps your certification current and takes only a few hours. It’s worth doing.

See more

The post DP-700 Certification Exam Preparation Guide appeared first on AzureOps.

In this post, I’ll walk you through how I built a Fabric Data Agent on top of the standard AdventureWorksDW dataset, and how you can too, even if you’re a complete beginner.

TL;DR

Microsoft Fabric’s Data Agent lets you ask questions about your data in natural language; no SQL, DAX, or visuals required. Think of it like a chat-based analyst that already knows your data model and never sleeps. This article walks you through building a Data Agent on top of AdventureWorksDW, writing agent instructions, adding data sources, and exploring advanced use cases like AI Foundry and multi-agent systems. By the end, you’ll be able to create your own interactive AI analyst in under 30 minutes.

Prerequisites:
1. You must have access to Microsoft Fabric Capacity specifically F2 or higher.
2. Only workspaces with Fabric-enabled capacity will show the “AI Agent” option in the “+ New” menu.
3.Your Fabric admin must enable the Data Agent preview in the Tenant Settings within the Microsoft Fabric Admin Portal. if you don’t have admin rights, contact your admin.

How to Create a Data Agent in Microsoft Fabric

If you’ve never done this before, don’t worry. Here’s a simple step-by-step guide.

Step 1: Open the Microsoft Fabric Workspace

1. Go to https://app.fabric.microsoft.com.
2. Select or create a new Workspace where your Lakehouse is located.

Step 2: Create a Lakehouse (if not already done)

1. Click New > Lakehouse and give it a name (e.g., demolakehouse).
2. Add your data using Shortcuts:
   • Go to the Tables pane inside the Lakehouse.
   • Click Add Shortcut > OneLake data.
   • Browse and select the relevant tables (e.g., from AdventureWorksDW).

I used the classic AdventureWorksDW dataset, the same one you’ve probably seen in BI demos. Instead of importing tables manually, I used OneLake Shortcuts, which connect existing data without duplicating it.

Step 3: Create the Data Agent

1. In your workspace, click New > Data agent (preview).
2. Choose the Lakehouse where your data is stored. You can also choose semantic model in case you have created semantic model on top of your lakehouse.
3. Give your agent a name (e.g., Sales Insights Agent).
4. Click Create — this opens the agent configuration screen.

Step 4: Add a Data Source to the Data Agent

Once the agent is created, you’ll need to give it instructions and connect it to your data.

1. Inside the agent editor, go to the Data Sources tab.
2. Click + Add data source and choose the Lakehouse as shown in the image below.

Step 5: Add the Agent Instructions

1. Go to the Agent Instructions tab.
2. Provide instuctions for the data agent about it’s goal and responsibilities.

Goal:
Assist users in analyzing sales, customers, geography, and product performance.

Key Relationships:
- FactInternetSales ➡ ProductKey ➡ DimProduct ➡ Subcategory ➡ Category
- FactInternetSales ➡ CustomerKey ➡ DimCustomer ➡ Geography
- FactInternetSales ➡ OrderDateKey ➡ DimDate

Ask me things like:
- Top-selling categories last year?
- Sales trend in Q1 2024?
- Which cities had highest average order size?

Answer in charts, tables, or summaries where helpful.

Step 6: Add the Data Source Instructions

1. Add the data source instruction to let data agent understand your data model.

yamlCopyEditname: AdventureWorksDW_Shortcut
type: Lakehouse
tables:
  - name: DimCustomer_shortcut
    description: Customer demographics and keys
  - name: DimDate_shortcut
    description: Calendar date dimension
  - name: DimProduct_shortcut
    description: Product attributes
  - name: DimProductCategory_shortcut
    description: Categories of products
  - name: DimProductSubcategory_shortcut
    description: Subcategories of products
  - name: DimGeography_shortcut
    description: Country, state, and city details
  - name: FactInternetSales_shortcut
    description: Sales facts including sales amount, quantity, and dates

Step 7: Test the Agent and fine tunning

Start asking questions to the agent like;

1. List top 10 cities by internet sales in last available 12 months.

2. List top 5 customers who made the highest purchases in the latest available year:

Notice that it also provides the SQL query it has executed on the lakehouse to get the required output.

If the data source and instructions are set correctly, the agent should respond with meaningful tables or summaries. In case you observe issues in the responses, you may need to fine tune the AI instructions and test again.

When Should You Use Data Agents?

If you’re already using:

• A well-modeled star schema
• Lakehouse tables or Delta format
• Power BI reports or semantic models
• Microsoft Fabric workspaces with clean relationships

Then a Data Agent will act like a bridge between your data and your business users. No technical barrier — just questions and answers.

Integrations and Advanced Use Cases

What makes Fabric Data Agents truly powerful is not just that they can answer questions — but that they can plug into larger AI and data ecosystems. Here are a few ways to take your agent to the next level.

1. Integration with Azure AI Foundry

Microsoft Azure’s AI Foundry allows you to orchestrate LLM-driven pipelines, combine ML models with data operations, and deploy custom copilots. Your Data Agent can be used here in multiple ways:

• As a ready knowledge base within Foundry apps
• To summarize and explain insights from streaming or batch pipelines
• As a modular AI skill within a larger orchestrated flow

For example, you could use the agent to automatically generate weekly sales summaries and send them via email to business users — no dashboards needed.

2. Use in Multi-Agent Systems with Copilot Studio

Microsoft recently announced integration of Fabric data agent with Copilot Studio. This means data agent also supports multi-agent patterns where different agents work together, each specializing in their role.

Your Data Agent can act as:

• A domain expert (e.g., Sales Agent, Inventory Agent)
• A knowledge base that other agents query
• A collaborative AI alongside others like a document reader, chatbot, or metric generator

Example: Multi-Agent Use Case

Imagine a scenario where you’ve built:

• A “Sales Insights Agent” (based on the Sales model)
• A “Customer Sentiment Agent” (that reads survey data or feedback documents)
• A “Decision Support Agent” (that aggregates insights from both)

Now, a user asks the Decision Support Agent:
“Should we invest more in red-colored bikes for Q3 promotions?”

Here’s what happens behind the scenes:

1. The Sales Agent looks up sales trends for red bikes from your Data Agent.
2. The Sentiment Agent checks recent customer feedback about product quality or color preferences.
3. The Decision Agent combines both and generates a recommendation — backed by data and context.

This makes your Data Agent a building block in a broader AI ecosystem, not just a standalone tool.

Final Thoughts

Data Agents in Fabric aren’t just experimental toys; they’re fully capable assistants for real business scenarios. Whether you’re a data engineer, analyst, or decision maker, you’ll find value in being able to converse with your data.

For me, this agent worked perfectly with the AdventureWorksDW model and started answering meaningful business questions within minutes, no DAX, no Power BI visuals, no joins to write.

This service is currently in preview. It does not currently render charts. Hopefully, future updates will enable it to generate Power BI level visuals on the fly.

See more

Pro tips:
1. Learn how to integrate Microsoft Copilot in SQL Server Management Studio 21.

The post Data Agent in Microsoft Fabric – Here’s How it Works appeared first on AzureOps.

In the example below, we want to drop and recreate a table ‘lh_gold.t_fact_day_sales’ in fabric lakehouse.

Error Message:

The typical error message we receive:

Fabric getting this error while dropping and recreating table| AnalysisException: [DELTA_CREATE_TABLE_WITH_NON_EMPTY_LOCATION] Cannot create table (‘lh_gold.t_fact_day_sales’). The associated location (‘abfss://7eaf3bc1-2d0c-43d9-bcbf-12b1195cb9a5@onelake.dfs.fabric.microsoft.com/cd923083-448c-43a9-bc86-94cfc95a7bc8/Tables/t_fact_day_sales’) is not empty and also not a Delta table.

Why Microsoft Fabric Can’t Auto-Delete Old Delta Files Like Databricks

Databricks has a built-in mechanism through its runtime to automatically manage files behind Delta Tables — especially when you DROP TABLE, TRUNCATE, or do a VACUUM. It deeply integrates with the storage layer (like DBFS, or direct Azure Data Lake) and uses dbutils and Delta Lake transaction logs (_delta_log/) to keep everything consistent.

However, Microsoft Fabric (as of 2024–2025):

• Uses a different storage abstraction: OneLake, not direct ADLS.
• Runs on Spark pools that don’t fully expose low-level file system utilities like dbutils.fs.
• Manages tables a little differently — especially when they’re shortcuts or external tables — and sometimes does not automatically delete orphaned files if a table is dropped.
• Also, Fabric’s Spark runtime is hardened and sandboxed for multi-tenancy and governance, which limits direct file system operations unless explicitly triggered by the user (via mssparkutils, Fabric UI, or API).

In short, Fabric prioritizes safety and governance over automatic cleanup, and doesn’t yet implement the same auto-vacuuming behaviors Databricks does.

Follow the steps below to delete Delta files from Fabric workspace programmatically.

1. Check if the Location Contains Files

from pyspark.sql.functions import col
files = mssparkutils.fs.ls(“abfss://<a href="mailto:7eaf3bc1-2d0c-43d9-bcbf-12b1195cb9a5@onelake.dfs.fabric.microsoft.com" target="_blank" rel="noreferrer noopener">7eaf3bc1–2d0c-43d9-bcbf-12b1195cb9a5@onelake.dfs.fabric.microsoft.com</a>/cd923083–448c-43a9-bc86–94cfc95a7bc8/Tables/t_fact_day_sales”)
display(files)

2. Delete Non-Delta Files (If Any)

mssparkutils.fs.rm(“abfss://<a href="mailto:7eaf3bc1-2d0c-43d9-bcbf-12b1195cb9a5@onelake.dfs.fabric.microsoft.com" rel="noreferrer noopener" target="_blank">7eaf3bc1–2d0c-43d9-bcbf-12b1195cb9a5@onelake.dfs.fabric.microsoft.com</a>/cd923083–448c-43a9-bc86–94cfc95a7bc8/Tables/t_fact_day_sales”, recurse=True)

3. Drop the Table Before Recreating

DROP TABLE IF EXISTS lh_gold.t_fact_day_sales;

4. Create the Table Again Using Delta Format

CREATE TABLE lh_gold.t_fact_day_sales
USING DELTA
LOCATION ‘abfss://7eaf3bc1–2d0c-43d9-bcbf-12b1195cb9a5@onelake.dfs.fabric.microsoft.com/cd923083–448c-43a9-bc86–94cfc95a7bc8/Tables/t_fact_day_sales’;

Alternative: Delete the Folder Manually

If you are unable to delete files programmatically, try these options:

• OneLake Explorer → Navigate to Tables/t_fact_day_sales and delete the folder.
• Azure Storage Explorer → Connect to the Fabric OneLake storage and manually delete the folder.

Once the folder is clean, retry the table creation.

Pro tips:
1. When using an external table, we can easily apply an alternative way to drop the folder. However, when working with a managed table, we have to use the fabric api as above.
2. If you want to automate deletion of old files from blob storage, follow this article.

See more

The post How to Delete Stale Delta Files in Microsoft Fabric Lakehouse appeared first on AzureOps.