DP-800 Exam Guide
SQL AI Developer Associate
What DP-800 actually tests, who it’s for, how it differs from DP-600 and DP-700, and how to prepare — covering Microsoft’s newest and first AI-focused SQL certification.
DP-800: Developing AI-Enabled Database Solutions is the exam for Microsoft’s Certified: SQL AI Developer Associate credential. It validates building AI-enabled database solutions — vector search, semantic search, and RAG workflows in T-SQL — across SQL Server 2025, Azure SQL, and SQL databases in Microsoft Fabric.
Unlike DP-600 and DP-700, which are scoped entirely to Fabric, DP-800 spans Microsoft’s whole SQL platform family.
DP-800 Exam Guide — Frequently Asked Questions
This guide reflects Microsoft’s published DP-800 skills-measured document and exam details as of July 2026. Microsoft periodically revises certification exams — always confirm current details at the official DP-800 study guide on Microsoft Learn before scheduling an exam. UIG Data Lab is independent and not affiliated with Microsoft.
What Makes DP-800 Different From Every Other Fabric Exam
DP-800 is not an AI certification that happens to touch SQL — it’s a SQL certification that covers AI. The exam assumes deep existing T-SQL knowledge and builds vector search, semantic search, and RAG capabilities directly on top of that foundation, rather than teaching AI concepts from scratch.
Why Microsoft Built DP-800
As enterprises build private AI applications, many are realizing that moving structured data out of secure, governed databases and into standalone vector stores creates real administrative and security overhead. DP-800 exists to validate the alternative: building AI capabilities — embeddings, vector search, RAG — directly inside SQL, where the data, governance, and security controls already live.
Microsoft announced the certification on its Skills Hub Blog on March 3, 2026, ran a discounted beta through April 3, 2026, and brought it to general availability in May 2026 — a fast timeline that reflects how central AI-in-database workloads have become to Microsoft’s data platform roadmap.
DP-800 at a Glance
| Detail | Value |
|---|---|
| Exam code | DP-800 |
| Exam name | Developing AI-Enabled Database Solutions |
| Certification | Microsoft Certified: SQL AI Developer Associate |
| Announced | March 3, 2026 |
| General availability | May 2026 |
| Duration | 100 minutes (120 minutes total seat time) |
| Passing score | 700 out of 1000 |
| Platforms covered | SQL Server 2025, Azure SQL, SQL databases in Microsoft Fabric |
| Delivery | Online proctored or Pearson VUE testing center |
Who DP-800 Is Actually For
DP-800 targets SQL developers, database administrators, and cloud solution architects — people who already understand indexes, query tuning, and T-SQL at a working level. It is not an entry point into either SQL or AI individually; it’s a bridge for SQL professionals moving into AI-enabled application work.
Every SQL Server DBA who already understands query performance, security, and modern table types has the right foundation. What DP-800 adds on top is the AI layer: calling language models from T-SQL, generating and searching vector embeddings, and building semantic search and RAG workflows without leaving the database engine.
DP-800 Skills Measured
Per Microsoft’s official skills-measured document, DP-800 is organized into three weighted domains.
| Domain | Weight | Key Topics |
|---|---|---|
| Design and develop database solutions | 35-40% | Tables, indexes, advanced T-SQL, JSON, graph queries, GitHub Copilot & MCP integration |
| Secure, optimize, and deploy database solutions | 35-40% | Encryption, Row-Level Security, query tuning, CI/CD with SQL Database Projects, Azure integration |
| Implement AI capabilities | 25-30% | Embeddings, vector search, hybrid search, RAG workflows using language models |
Underneath those three headline domains, published study guides consistently list a specific technical surface: temporal tables, ledger tables, graph tables, and in-memory OLTP under solution design; Always Encrypted, Dynamic Data Masking, and Managed Identity under security; and vector data types, VECTOR_DISTANCE, DiskANN indexes, and external models under the AI domain.
The AI Domain, in Practical Terms
The “Implement AI capabilities” domain is the part every other Microsoft SQL certification has skipped until now. It covers generating vector embeddings from text directly in T-SQL, storing them using native vector data types, and running similarity search with functions like VECTOR_DISTANCE against DiskANN-backed indexes for fast approximate nearest-neighbor lookups at scale.
On top of that, it covers hybrid search — combining traditional keyword search with vector similarity — and full RAG (Retrieval-Augmented Generation) workflows, where a query retrieves grounding context from the database before an external language model generates a response. The goal is grounding LLM outputs in real data and reducing hallucination, using infrastructure a SQL professional already knows how to secure and scale.
DP-800 vs DP-600 vs DP-700
| Exam | Certification | Scope |
|---|---|---|
| DP-600 | Fabric Analytics Engineer Associate | Microsoft Fabric only — analytics engineering, semantic models, Power BI |
| DP-700 | Fabric Data Engineer Associate | Microsoft Fabric only — data engineering, pipelines, Spark, real-time |
| DP-800 | SQL AI Developer Associate | SQL Server 2025, Azure SQL, and Fabric SQL databases — AI-enabled T-SQL development |
DP-600 and DP-700 both live entirely inside Microsoft Fabric. DP-800 is structurally different — it’s the only one of the three that isn’t Fabric-exclusive, since the same AI-in-SQL skills apply whether the database is a Fabric SQL database, Azure SQL, or an on-premises SQL Server 2025 instance.
For a data engineer already holding DP-700, DP-800 is a natural complement rather than a replacement — it validates a different, adjacent skill set focused specifically on the AI layer inside SQL rather than pipeline and Spark engineering.
How to Prepare for DP-800
- Start with the official study guide. Microsoft Learn’s DP-800 study guide is the source of truth for the current skills-measured document — features get added as the underlying products evolve, so check it close to your exam date rather than relying on an older summary.
- Work through the self-paced Microsoft Learn plan, “Become a SQL AI Developer: Prepare for Certification Exam DP-800,” which maps directly to the three scored domains.
- Practice vector and semantic search hands-on in Azure SQL Database rather than only reading about it — DiskANN indexing and VECTOR_DISTANCE behavior are easier to internalize by running real queries than by memorizing syntax.
- Use the exam sandbox Microsoft provides before test day to get comfortable with the online proctored environment itself, separate from the content.
- Check for active discount promotions before registering — Microsoft Learn Cloud Skills Challenges around Build or Ignite, employer Microsoft Partner Network vouchers, and Fabric-focused virtual training days have all offered free or discounted DP-800 registration since launch.
One planning note: DP-800 launched in beta, and beta exam scoring is delayed — results aren’t released immediately, and a practice assessment may not exist yet for a brand-new exam. Now that DP-800 is GA as of May 2026, standard scoring and practice assessment availability should apply, but always confirm current status on the official exam page before scheduling.
Where DP-800 Fits in a Broader Fabric Career Path
DP-800 sits alongside DP-600 and DP-700 as a third pillar of Microsoft’s data platform certification track, but it’s the one most directly tied to where the industry is heading — AI as a first-class database workload rather than a bolt-on service.
For data professionals already working inside Microsoft Fabric day to day, understanding the F-SKU and licensing model those other two certifications assume is worth having alongside DP-800’s AI-specific skill set — the Fabric Pricing Calculator and the Fabric vs Power BI guide both cover that platform context in detail.