Quiz · 4 questions

🐘 pgvector and the SQL Hybrid

When your vectors should live next to your relational data

Level 0Scout

0 XP0/41 lessons0/10 achievements

0/120 XP to next level120 XP to go0% complete

Quiz

01Your team already runs Postgres for the relational app. Most retrieval results need to join with 3 other tables for the LLM prompt. Which choice usually wins?

02What happens if you create a vector(1024) column and then later switch to a 1536-dim embedding model?

03An HNSW index returns approximate results. What is 'recall' in this context?

04You write a hybrid query with WHERE filters plus an ORDER BY embedding <=> $1, and only 3 of the requested 10 results survive the filter. What is the standard fix?

← PreviousMigrations, Re-embedding, and Zero Downtime 🔀 Hybrid Search and Reranking →BM25 Isn't Dead — Pair It With Vectors

Spotted a bug or have feedback on this page?Report an Issue

Comments 0

🔔 Reply notifications (sign in)

No comments yet — be the first.