A video recommendation system stores 200 million video embeddings in Azure SQL Database. The team needs sub-50ms query latency. They tested HNSW and IVFFlat indexes. HNSW gives 95% recall@10 at 45ms, while IVFFlat gives 88% recall@10 at 20ms. The product team requires at least 90% recall. What should the team do?

Use HNSW because it meets the recall requirement and the 45ms latency is within the 50ms budget.

Use IVFFlat and increase the number of lists to improve recall to 90% while keeping latency under 50ms.

Use IVFFlat because it is faster, even though recall is below the requirement.

Use HNSW and increase ef_search to improve recall beyond 95%, accepting higher latency.

When designing a vector column, what factor is most important to consider regarding vector size?

The dimensionality must match the embedding model's output

The size should be a power of two for performance

The size is determined by the number of rows in the table

The size should be as small as possible to save space

A search team is tuning an HNSW vector index in Azure SQL Database. They observe that increasing ef_search from 100 to 200 improves recall@10 from 85% to 92% but increases p95 latency from 30ms to 60ms. The application requires p95 latency under 50ms. What should they do?

Keep ef_search at 100 because latency is more important than recall.

Keep ef_search at 200 because recall is more important than latency.

Set ef_search to 150 and measure the recall and latency to find a point that meets both requirements.

Switch to IVFFlat, which will always meet both requirements.

Microsoft Certified:SQL AI Developer Associate

Domain 3Objective 2

Design and Implement Intelligent Search DP-800 Practice Questions (Page 7)

Part of the Implement AI capabilities in database solutions domain, which accounts for 25–30% of the DP-800 exam. Microsoft does not publish an official question count, but from its 120-minute exam (~50–80 total, ~13–24 in this domain), expect 4–8 from this objective — we provide 47 practice questions to prepare you well beyond it. (estimate)

47questions here

10free pages

10concepts

25–30%of the exam

Questions 31–35

31expert · hard
A video recommendation system stores 200 million video embeddings in Azure SQL Database. The team needs sub-50ms query latency. They tested HNSW and IVFFlat indexes. HNSW gives 95% recall@10 at 45ms, while IVFFlat gives 88% recall@10 at 20ms. The product team requires at least 90% recall. What should the team do?
Select an answer first
32foundation · easy
When designing a vector column, what factor is most important to consider regarding vector size?
Select an answer first
33foundation · easy
Which Transact-SQL predicate is used in a WHERE clause to perform a full-text search for words or phrases in a table?
Select an answer first
34expert · hard
A search team is tuning an HNSW vector index in Azure SQL Database. They observe that increasing ef_search from 100 to 200 improves recall@10 from 85% to 92% but increases p95 latency from 30ms to 60ms. The application requires p95 latency under 50ms. What should they do?
Select an answer first
35foundation · easy
Which vector index type in SQL Server 2025 is designed for Approximate Nearest Neighbor (ANN) search?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to DP-800

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by Microsoft. “DP-800” is a trademark of its owner, used for identification only.

Design and Implement Intelligent Search DP-800 Practice Questions (Page 7)

Questions 31–35

When designing a vector column, what factor is most important to consider regarding vector size?

Which Transact-SQL predicate is used in a WHERE clause to perform a full-text search for words or phrases in a table?

A search team is tuning an HNSW vector index in Azure SQL Database. They observe that increasing ef_search from 100 to 200 improves recall@10 from 85% to 92% but increases p95 latency from 30ms to 60ms. The application requires p95 latency under 50ms. What should they do?

Which vector index type in SQL Server 2025 is designed for Approximate Nearest Neighbor (ANN) search?

Keep studying — Implement AI capabilities in database solutions