Develop AI Solutions by Using Azure Database for PostgreSQL AI-200 Practice Questions (Page 3)

Questions 11–15

1. 11application · medium

   A company runs a vector search workload on Azure Database for PostgreSQL using pgvector. The workload has a large in-memory working set of embeddings and experiences high CPU usage during searches. The team wants to improve performance without changing the application code. Which configuration change should they make?

   • AScale up the server to a higher compute tier with more vCores and memory.
   • BChange the PostgreSQL version to a newer release.
   • CSet max_connections to a higher value to allow more concurrent queries.
   • DIncrease the storage size to allow more disk space for the table.
   Reveal Solution Discussion Select an answer first
2. 12expert · medium

   A company is running a pgvector workload on Azure Database for PostgreSQL. The database is experiencing high CPU usage during similarity searches. The team has already created an HNSW index. Which additional action should they take to reduce compute overhead?

   • ADisable the HNSW index and use sequential scans.
   • BIncrease the ef_search parameter to improve recall.
   • CIncrease the shared_buffers parameter to cache more data.
   • DReduce the ef_search parameter to lower the number of candidates examined during search.
   Reveal Solution Discussion Select an answer first
4. 13expert · medium

   A developer is implementing a semantic search feature that must return results within 200 milliseconds. The database contains 5 million embeddings. The developer has created an HNSW index, but queries are still slow. Which optimization should the developer apply?

   • AIncrease the number of lists in an IVFFlat index.
   • BUse a sequential scan instead of the index.
   • CIncrease the ef_search parameter to improve recall.
   • DReduce the ef_search parameter to speed up queries at the cost of some recall.
   Reveal Solution Discussion Select an answer first
5. 14foundation · easy

   Which PostgreSQL extension is required to store and query vector embeddings in Azure Database for PostgreSQL?

   • Apg_trgm
   • Bpgvector
   • Cuuid-ossp
   • DPostGIS
   Reveal Solution Discussion Select an answer first
6. 15foundation · easy

   Which SQL approach is used in a RAG pattern to combine a metadata filter with a vector similarity search in PostgreSQL?

   • ASELECT * FROM documents WHERE embedding_column = query_vector
   • BSELECT * FROM documents ORDER BY metadata_column
   • CSELECT * FROM documents GROUP BY metadata_column
   • DWHERE metadata_column = value AND embedding_column <-> query_vector ORDER BY distance LIMIT n
   Reveal Solution Discussion Select an answer first