Microsoft Certifications
Microsoft Data Engineer Certification
DP-203 retired March 31, 2025. DP-700 is now the Microsoft data engineer cert, and it is a meaningfully different exam from the one your senior teammates took. This guide covers the retirement, what DP-700 actually tests, the F-SKU pricing trap, how it compares to AWS and Google, and an eight-week study plan that maps directly to exam objectives.
Updated May 2026·By The DataDriven Team
What this guide actually says
- 01DP-203 retired March 31, 2025. DP-700 is the current Microsoft data engineer cert.
- 02DP-700 tests Fabric: Lakehouse, Warehouse, and Real-Time Intelligence. Synapse-only knowledge is no longer enough.
- 03If your target shop is on classic Azure (ADF + Synapse), DP-203 retirement still hurts because there is no like-for-like replacement.
- 04DP-203 holders should renew through the Fabric learning path before their renewal window closes.
- 05DP-900 is the prereq for career switchers. Skipping it usually means failing DP-700 once and re-paying.
- 06F-SKU pricing is a graded topic. Memorize the F2 / F8 / F64 anchor points before exam day.
By the numbers
The current state of the Microsoft data engineer cert track in 2026.
2025
DP-203 retired
DP-700
Current DE cert
$165
Exam fee
120 min
Time limit
What changed in March 2025 (and what hiring managers know)
DP-203 retired. DP-700 took over. The reasoning behind the move is not random, and senior interviewers read the difference cleanly.
Microsoft retired DP-203 because the company bet hard on Fabric and needed the cert track to follow. The exam that taught Synapse Dedicated SQL Pools and PolyBase was actively confusing for new candidates joining Fabric-first projects. DP-700 replaces it with a syllabus calibrated to where Microsoft customers are actually shipping.
The political reasoning
Fabric consolidates Synapse, Azure Data Factory, Power BI, and Real-Time Intelligence under a single SaaS SKU. Microsoft sells one capacity, one bill, one security model. Retiring DP-203 forced the cert market to follow the product roadmap.
The customer reality
Most enterprises in 2026 still run a hybrid: some Synapse, some Fabric, occasionally Databricks-on-Azure for ML. Greenfield projects start in Fabric. Existing Synapse Dedicated Pools rarely migrate cleanly. DP-700 is calibrated for the greenfield shape, which is why DP-203 holders feel the gap.
The hiring manager view
A 2024 DP-700 holder signals modern stack fluency. A 2023 DP-203 holder signals legacy Synapse experience, which is still valuable but harder to deploy on new projects. Recruiters and hiring managers in 2026 read these as different signals, not interchangeable ones.
Watch out
If you are studying from a 2023 or 2024 DP-203 prep book, throw it out. The Synapse Dedicated Pool distributions section, the PolyBase external table syntax, and the legacy Mapping Data Flows material are not on DP-700 and will eat your study time.
DP-203 retirement timeline
The dates senior interviewers reference when they read your resume.
| Date | Event | What it meant |
|---|---|---|
| November 2023 | DP-700 announced | Microsoft signals Fabric as the strategic surface. Most candidates miss the announcement. |
| Q1 2024 | DP-700 enters beta | Beta testers report a meaningfully different exam shape with KQL and Eventstreams as new sections. |
| Q2 2024 | DP-700 GA | Exam launches at standard $165 price. DP-203 still active and still booked heavily. |
| Late 2024 | Microsoft Learn DP-203 path frozen | Updates stop. The signal is clear: DP-203 is in maintenance mode. |
| March 31, 2025 | DP-203 retires | Final exam date. New registrations end. Existing certifications remain valid through their renewal cycle. |
| 2026 onward | DP-700 only | Microsoft data engineer associate track is single-cert. Renewals route through the Fabric learning path. |
DP-700 in detail: what is actually on the exam
Six workload areas. The boundaries between them are the source of every scenario question.
Lakehouse
Fabric Lakehouse
Delta tables backed by OneLake. T-SQL endpoint for ad hoc reads, Spark notebooks for transformation. Shortcuts let you mount data from another workspace, an ADLS Gen2 account, or an S3 bucket without copying bytes. The exam tests whether you understand that a shortcut is a metadata pointer, not a replication.
Warehouse
Fabric Warehouse
T-SQL surface that looks like Synapse but is not Synapse. The semantic differences matter on the exam. Identity columns, schema-bound views, and cross-database queries behave differently from Dedicated SQL Pools. Storage still lives in OneLake as Delta, but the query engine is the new Polaris-derived MPP, not the legacy Synapse pool.
Pipelines
Data pipelines and Dataflow Gen2
Fabric Data pipelines are the lift-and-shift of Azure Data Factory. Dataflow Gen2 is the Power Query authoring surface for low-code transformation. Pipelines win for orchestration, parametrization, and copy-at-scale. Dataflows win for analyst-authored cleanup. Picking the wrong tool is one of the most common scenario-question traps.
Real-Time
Real-Time Intelligence
Eventstreams ingest from Event Hubs, Kafka, IoT Hub, or HTTP. Eventhouses store the data in KQL databases (the Kusto engine that powers Application Insights and Azure Data Explorer). KQL is the differentiator most candidates skip. Plan to learn summarize, mv-expand, make-series, and the let bindings that show up in every scenario question.
Lifecycle
Deployment, security, capacity
Deployment pipelines move artifacts from dev to test to prod workspaces. Git integration backs everything with version control (Azure DevOps or GitHub). OneLake security uses workspace roles plus item-level permissions, with row and column security inherited from the underlying Delta table. Capacity is the F-SKU you paid for. If a workload bursts past it, requests throttle.
Semantic
Semantic models and Direct Lake
Direct Lake mode reads Delta files in OneLake without import or DirectQuery. It is fast for Power BI, but it inherits the limits of Vertipaq under the hood. The exam asks scenario questions about when to fall back to import or DirectQuery, and how to diagnose a Direct Lake fallback in the capacity metrics app.
The Real-Time Intelligence section is where most candidates lose points. Eventstreams, Eventhouses, and KQL databases are net new material that DP-203 holders never saw. The exam will give you a scenario about high-cardinality clickstream or IoT telemetry and ask you to design the ingest path and the query layer in the same answer.
ArchitecturePractice the streaming pattern
Two Hundred Million Redirects
Billions of clicks. One tiny code. Two very different clocks.
Cost: the F-SKU question
Fabric is sold by capacity. The exam grades capacity sizing more aggressively than candidates expect, and most prep books skim it.
Capacity is denominated in F-SKUs, where the number is the count of capacity units. Capacity smoothing and bursting let workloads briefly exceed the SKU, then throttle when the smoothing window fills. Pricing is hourly and most customers either reserve capacity for a year (about 40 percent discount) or run pay-as-you-go for dev/test workspaces.
| F-SKU | Hourly | Monthly | Typical use |
|---|---|---|---|
| F2 | $0.36 / hour | $262 / month | Toy. Demos and sandboxes only. |
| F4 | $0.72 / hour | $525 / month | Solo developer. Tight. |
| F8 | $1.45 / hour | $1,057 / month | Small team or single product line. |
| F16 | $2.90 / hour | $2,114 / month | Mid-size analytics team. |
| F32 | $5.81 / hour | $4,242 / month | Multi-team workspace. |
| F64 | $11.62 / hour | $8,481 / month | Mid-enterprise. Power BI Pro included for all viewers. |
| F128 | $23.23 / hour | $16,956 / month | Enterprise. Multiple capacities common. |
Exam trap
F64 is the magic line. At F64 and above, every viewer in the tenant gets Power BI Pro entitlement included. Below F64, non-author viewers still need their own Pro license. Scenario questions test this directly: a 200-viewer tenant on F32 looks cheaper on paper but costs more after Pro licenses than a single F64 capacity.
“Microsoft retired DP-203 because they want you on Fabric. Hiring managers want you to ship. Both are true.”
DP-700 vs the other clouds
A side-by-side that maps DP-700 against the other certs you might be choosing between in 2026.
| Exam | Difficulty | Scope | Reach (2026) | Hiring signal | Transferability |
|---|---|---|---|---|---|
| Microsoft DP-700 | Medium-Hard | Fabric Lakehouse, Warehouse, Real-Time, Pipelines | Strongest in regulated enterprises (finance, healthcare, gov) | Strong inside the Microsoft ecosystem, modest outside | Limited. F-SKU and OneLake concepts do not map to other clouds. |
| AWS DEA-C01 | Medium | Glue, Redshift, Kinesis, Lake Formation, S3 | Broadest cloud DE market share in 2026 | Strong almost everywhere. Default cert when undecided. | High. Most patterns transfer to Azure and GCP equivalents. |
| Google Pro Data Engineer | Hard | BigQuery, Dataflow (Beam), Pub/Sub, Bigtable, Vertex AI | Smaller footprint, concentrated at GCP-first shops | Highest per-cert prestige. Hard to fake. | High for streaming, ML pipelines, watermarking concepts. |
| Databricks DEA | Medium | Delta Lake, Spark, medallion, Unity Catalog | Hot. Lakehouse adoption accelerating across clouds. | Strong for any company running Spark, regardless of cloud | High. Spark + Delta knowledge applies on AWS, Azure, GCP. |
What interviewers grade on at Microsoft-stack shops
Real questions from Fabric-shop interview loops in 2026. The patterns recur.
Q01
Walk me through your Fabric workspace organization for a multi-domain analytics platform
Strong answers separate workspaces by domain (sales, finance, supply chain) and lifecycle (dev, test, prod), then explain how OneLake shortcuts let teams share canonical Gold tables without copying. Mention deployment pipelines, capacity assignment, and the trade-off between one large F-SKU and many smaller capacities. Weak answers describe a single 'analytics' workspace and miss the governance question entirely.
Q02
When would you pick a Fabric Lakehouse vs Warehouse vs Eventhouse?
Lakehouse for raw ingestion, Spark transformation, and ML feature engineering. Warehouse for T-SQL workloads where analysts expect SQL Server semantics and stored procedures. Eventhouse for high-cardinality time-series and log-style data where you need sub-second KQL queries over billions of rows. The interviewer is checking whether you understand that all three sit on OneLake but use different engines.
Q03
OneLake shortcuts: explain the security implications when shortcutting across workspaces
A shortcut inherits the source table's row-level security and column masking, but the destination workspace's roles control who can resolve the shortcut. That gap is where leaks happen. Strong answers also mention that shortcuts to external storage (ADLS Gen2, S3) authenticate using the source connection, not the destination, which can route reads through unintended identities.
Q04
Your Eventstream is dropping events under load. Diagnose
Walk through the layers. First, capacity throttling at the Eventhouse (check capacity metrics app for throttled requests). Second, Eventstream throughput unit limits. Third, the source side: are Event Hubs partitions saturated, or is the producer batching badly? Strong answers reference the Eventstream monitoring view and the difference between dropped events and rejected events.
Q05
Design a CDC pipeline from on-prem SQL Server into Fabric
Most candidates start with Data Factory's self-hosted integration runtime. Better answers consider a SQL Server CDC enable + Debezium-to-Event-Hubs path, then Eventstream into a Bronze Lakehouse Delta table, then a notebook merging into Silver. The interviewer wants to see that you understand initial snapshot vs ongoing delta, idempotency on retry, and how to handle schema drift on the source side.
OneLake security: the part candidates underprepare
The four layers that determine who sees what in Fabric. The exam tests the gaps between them.
OneLake security is layered. Workspace roles set the ceiling. Item-level permissions narrow it. Row-level security travels with the Delta table. Shortcuts inherit one side and require the other. Most candidates know one of the four well and lose points on the others.
Workspace
Workspace roles control item access
Admin, Member, Contributor, Viewer. The four roles control who can see and modify items in the workspace. Viewers see lakehouses and warehouses but cannot edit them. Contributors can create new items. Member adds the right to manage workspace settings. Admin owns the workspace and assigns roles.
Item
Item-level permissions override workspace roles down
You can grant a user read on a single Lakehouse without giving them the rest of the workspace. The exam tests scenarios where a Viewer needs read on Gold tables but no access to Bronze. Item permissions answer this. They cannot escalate above the workspace role, only restrict beneath it.
Row
Row-level security travels with the Delta table
RLS defined on a Lakehouse Delta table is enforced uniformly: T-SQL queries through the SQL endpoint, Spark notebooks reading the Delta, and shortcuts pointing at the table all see the filter. This is the cleanest part of the OneLake security story and the most-tested.
Shortcut
Shortcuts inherit source security but consumer authorization
When you shortcut a table from Workspace A into Workspace B, the source table's RLS and column masking still apply. But the act of resolving the shortcut requires permissions in the destination workspace. Misconfigure either side and you either over-share data or break a published dashboard.
Interview tip
When asked about Fabric security, walk all four layers in order even if the question is about one. Senior interviewers grade structural answers higher than narrow ones. Saying "workspace roles, then item permissions, then RLS, then shortcuts" before diving in shows you understand the model.
Interview soundbites
Short, defensible answers to the questions that recur in Microsoft-stack DE interviews. Memorize the structure, not the words.
Lakehouse vs Warehouse
When asked
Lakehouse first if the workload is Spark, ML feature engineering, or open-format storage you need to share with non-Microsoft consumers. Warehouse first if the workload is T-SQL with stored procedures, the team is composed of SQL developers, and you need full ANSI semantics for joins and window functions.
Direct Lake fallback
When asked
Direct Lake reads Delta files in OneLake without import or DirectQuery. It falls back to DirectQuery when the table exceeds Vertipaq limits, when calculated columns block the lake path, or when the user lacks the proper SQL endpoint permissions. Diagnose with the Capacity Metrics app's fallback indicator.
Capacity throttling
When asked
Fabric smoothes capacity over a 24-hour window. Workloads can burst above the SKU briefly, then throttle when the smoothing window fills. The capacity metrics app shows pending requests and throttle minutes. Right answer to a throttle question is rarely 'increase the SKU'. It is usually 'right-size the workload, schedule heavy jobs off-peak, or move the noisy item to its own capacity'.
Eventstream durability
When asked
Eventstreams are not durable storage. They route events. Durability lives at the destination: an Eventhouse, a Lakehouse, or a Custom App with retry logic. Treat the Eventstream like a Kafka Streams topology, not like Kafka itself. The exam asks this exact distinction in scenario form.
Schema drift
When asked
Spark notebooks handle schema drift natively with mergeSchema = true on Delta writes. Pipelines and Dataflow Gen2 do not, and they fail loudly when the source adds a column. Strong answers walk through both paths and recommend Spark notebooks for sources where schema drift is common.
Cross-cloud
When asked
Fabric can shortcut to ADLS Gen2 and to S3, but not to GCS as of mid-2026. The shortcut authenticates through the source connection, so a single workspace can read tables in S3 without the data ever copying into OneLake. This is the answer to the 'we have a Snowflake bill on AWS, can we keep the data there?' question.
Myth vs reality
Five things candidates believe walking into DP-700 prep. Hiring managers and senior interviewers do not.
The Myth
Fabric replaces Synapse.
The Reality
Synapse Dedicated SQL Pools are still GA and supported. Most large customers run both for years during migration. The exam expects you to know the difference and choose between them.
The Myth
DP-700 is just DP-203 with Fabric chapters bolted on.
The Reality
DP-700 is a meaningfully different exam. KQL and Eventstreams are full sections. Synapse-specific topics (dedicated pool distributions, PolyBase) are gone. Studying DP-203 material will leave you 30 percent under-prepared.
The Myth
Fabric is just Power BI dressed up.
The Reality
At the data engineering layer, Fabric runs Spark, Delta, and the Kusto engine. None of that is Power BI. The semantic model layer touches Power BI, but DP-700 grades the engineering tier independently.
The Myth
Microsoft's Azure data engineer market shrank when DP-203 retired.
The Reality
It grew. Regulated enterprises (finance, healthcare, government) accelerated Fabric adoption in 2025 and 2026 because the unified billing and OneLake security model fit their compliance posture.
The Myth
I can use my AWS knowledge to pass DP-700.
The Reality
The F-SKU capacity model and Fabric workspace concepts have no AWS analogue. Plan to study the pricing layer and the OneLake security model from scratch even if you are senior on AWS.
Decision matrix
Use this if you have ten seconds. The answer that fits your situation is one row away.
If your situation is
Pick
Why
Targeting Microsoft enterprise shops
DP-700
Direct match for the platform they actually run.
Already work in Synapse, want renewal path
DP-700
DP-203 retired. DP-700 is the official Synapse-to-Fabric bridge.
Power BI developer pivoting to DE
DP-900 then DP-700
DP-900 builds the data vocabulary you need before DP-700 hits.
Multi-cloud consultant
AWS DEA-C01 first, DP-700 second
AWS gives you the broader market. DP-700 covers the Microsoft engagements.
Pure data engineer at AWS-only shop
Skip DP-700, take AWS DEA-C01
DP-700 will not move the needle if your stack never touches Azure.
ML engineer needing Microsoft credentials
AI-102 instead
AI-102 (Azure AI Engineer) maps to your work. DP-700 will not.
Career switcher, no cloud background
DP-900 first, then DP-700
Skipping fundamentals usually means failing DP-700 once and re-paying.
How to study for DP-700 in 8 weeks
A week-by-week plan calibrated to the actual exam blueprint, not the marketing copy.
- 01
Week 1 to 2: Microsoft Learn DP-700 path
Complete the official DP-700 learning path on Microsoft Learn (free). Spin up a Fabric trial tenant and confirm you can create a workspace, a Lakehouse, and a Warehouse. The trial includes 60 days of full F-SKU capacity, which is enough for the entire study cycle. Do not skip the labs. The exam scenario questions assume hands-on familiarity with the workspace UI. - 02
Week 3: Build a medallion pipeline using OneLake shortcuts
Ingest a real public dataset (NYC taxi, GitHub archive, anything reasonably large) into a Bronze Lakehouse, transform it with a Spark notebook into Silver, then aggregate into a Gold Warehouse table. Use a OneLake shortcut to expose Gold to a second workspace as if a downstream team consumed it. This is the single highest-ROI hands-on exercise for the exam. - 03
Week 4: Eventstream and Eventhouse hands-on
Stand up an Eventstream from a sample data source (the built-in Bicycles or Stocks generator works). Land it in an Eventhouse. Write KQL queries using summarize, bin, mv-expand, and make-series. Most candidates underprepare here and lose 15 to 20 percent of their score. Do not skip the windowing functions in KQL. - 04
Week 5: Practice exams (MeasureUp and Whizlabs)
Take a full timed practice exam on MeasureUp or Whizlabs. Score yourself honestly. For every question you get wrong, write a one-paragraph explanation of why the right answer is right and why the others are wrong. This 'why-not' analysis catches the conceptual gaps a passing score on flash cards hides. - 05
Week 6: Cost and capacity scenarios
DP-700 grades F-SKU sizing scenarios more aggressively than candidates expect. Memorize the F2 / F8 / F64 anchor prices. Understand how capacity smoothing, bursting, and throttling work. Practice questions where the answer is 'pick a smaller F-SKU and turn on autoscale' versus 'pick a larger F-SKU and dedicate it'. - 06
Week 7: Deployment pipelines and Git integration
Configure Git integration on a workspace. Make a change, push it, deploy to a staging workspace via a deployment pipeline. Understand the difference between selective deployment, deployment rules, and parameter overrides. The exam includes at least one scenario question about promoting a parameterized pipeline through dev / test / prod. - 07
Week 8: Final timed practice exam
Take a final timed practice exam in one sitting under exam-day conditions. No notes. No pausing. If you score above 80 percent, schedule the real exam within seven days. If you score below 70 percent, do not book it yet. Re-do the weakest section's hands-on labs and re-test before scheduling.
Hands-on artifacts to build before exam day
If you can build all five of these end to end without notes, you are ready.
- A medallion Lakehouse with Bronze, Silver, Gold layers wired up by Spark notebooks.
- A Warehouse view that joins Gold Lakehouse Delta tables via T-SQL.
- A Fabric Data pipeline that orchestrates the Spark notebooks, with parameters for environment.
- An Eventstream from the built-in Bicycles generator into an Eventhouse, with a KQL summarize query.
- A deployment pipeline that promotes all of the above from a dev to a test workspace.
Common pitfalls
Patterns that appear in failed first attempts at DP-700. Avoid these and your second sitting becomes your only sitting.
Pitfall
Studying DP-203 material and assuming it covers DP-700
About 30 percent of DP-700 is net new content. Studying old material gives a false sense of preparation. Throw out the 2023 prep books and start from the current Microsoft Learn DP-700 path.
Pitfall
Skipping KQL because 'I am not a streaming engineer'
KQL is on the exam regardless of your role. The Real-Time Intelligence section is roughly 20 percent of the score. You will not pass without basic KQL fluency: summarize, bin, where, project, mv-expand.
Pitfall
Memorizing F-SKU prices but not the F64 license boundary
The exam asks scenario questions about Power BI Pro licensing. F64 is the line where viewer Pro licenses are included in the capacity. Below F64, you still pay per-viewer. Candidates who memorize prices but miss this fail the licensing scenario every time.
Pitfall
Treating Direct Lake like DirectQuery
Direct Lake is a different mode with different limits. Calculated columns, calculated tables, and certain DAX patterns force a fallback. The exam grades whether you know when Direct Lake works and when you have to fall back, not just that the mode exists.
Pitfall
Ignoring deployment pipelines and Git integration
Several scenario questions assume you have promoted artifacts from dev to test to prod. If you have only ever worked in a single workspace, you will guess wrong on the deployment rule and parameter override questions. Practice the flow once end to end before the exam.
Practice the data patterns Microsoft-stack interviewers test
The cert validates your knowledge. These problems validate your skills.
SQLOpen
The Duplicate Detection Sprint
Same email, different rows. Spot the repeats.
Data ModelingOpen
The Customer Who Changed
She moved. She upgraded. She became someone new. The record has to keep up.
ArchitectureOpen
Two Hundred Million Redirects
Billions of clicks. One tiny code. Two very different clocks.
SparkOpen
Fix Skewed Viewing Events Pipeline
Your nightly Spark job just paged you. One task has 40% of the data.
Frequently asked questions
Is DP-203 still worth taking in 2026?+
No. DP-203 retired on March 31, 2025. New candidates cannot register for it. Existing holders are valid for the remaining renewal window but cannot retake. DP-700 is the current Microsoft data engineer associate cert.
What happens to my DP-203 cert if I already have it?+
It remains valid until your renewal date. Microsoft offers a transitional renewal path through the Fabric learning content on Microsoft Learn. You should plan to renew through DP-700 study material since your existing renewal is the last one tied to the DP-203 lineage.
How hard is DP-700 compared to DP-203?+
Comparable difficulty, different scope. DP-700 trades the deep Synapse Dedicated Pool material for KQL, Eventstreams, OneLake security, and F-SKU capacity sizing. Candidates strong on classic Azure data services tend to underestimate the KQL and Real-Time Intelligence sections.
Do I need DP-900 before DP-700?+
Not formally, but yes in practice if you are new to data concepts. DP-900 is two weeks of study and gives you the vocabulary (relational vs non-relational, batch vs stream, fact vs dimension) that DP-700 assumes. Career switchers who skip DP-900 fail DP-700 more often than the cohort that took it.
How much does Microsoft Fabric cost in production?+
Fabric is sold by capacity. F2 starts at $0.36 / hour (about $262 / month). F64 is $11.62 / hour (about $8,481 / month) and is the smallest SKU that includes Power BI Pro for all viewers. Most mid-market customers land on F32 or F64. Enterprise tenants run multiple capacities to isolate workloads.
Does DP-700 expire?+
Yes. Microsoft role-based associate certs (including DP-700) require renewal once a year. The renewal is a free open-book online assessment, takes about 30 minutes, and covers features added since your last renewal.
Is KQL really on the exam, even for non-streaming roles?+
Yes. The Real-Time Intelligence section is roughly 20 percent of DP-700, and KQL questions appear within it. You do not need to be expert level, but you should be able to read a KQL query, identify what summarize and bin do, and pick the right time-series operator for a scenario.
Should I pair DP-700 with another cloud cert?+
If your work is multi-cloud, yes. AWS DEA-C01 is the most common pairing because most companies that use Fabric also run something on AWS for legacy reasons. Avoid stacking three associate-level cloud certs at once. The marginal hiring signal drops sharply after the second.
The cert proves what you know. Practice proves what you can ship.
DataDriven covers SQL, Python, system design, data modeling, and Spark at interview difficulty. Match the cert with the muscle memory.