Find Any Work in Seconds — With AI That Understands Context

From keyword precision to semantic understanding, Artclerk search adapts to how curators actually think and ask questions.

Capabilities

Search That Thinks Like a Curator

PostgreSQL tsvector with GIN index across titles, stock numbers, notes, and more — instant results, even across large collections.

OpenAI text-embedding-3-small embeddings with pgvector HNSW index. Hybrid ranking blends 60% keyword and 40% semantic relevance.

Filter by status, artist, medium, year, and more. Removable filter pills with debounced input for a fluid search experience.

Ask natural-language questions about your collection — "show me all oil paintings from the 1960s in need of conservation review."

Under the Hood

Artclerk combines the precision of keyword search with the intelligence of vector similarity.

Final Score = 0.6 × keyword_rank + 0.4 × semantic_similarity

Keyword search uses PostgreSQL full-text with GIN index. Semantic search uses pgvector HNSW index with OpenAI embeddings.