Skip to main content

Keyword/faceted search for documents

POST 
/api/v1/search

Keyword search endpoint - Search documents using Meilisearch for full-text and faceted search.

This endpoint provides fast keyword search with typo tolerance, relevance ranking, and powerful filtering capabilities.

Features:

  • Full-text search across title, description, abstract, text content
  • Typo tolerance (finds "machnie" when searching for "machine")
  • Relevance ranking based on TF-IDF and proximity
  • Faceted filtering by dataset, status, MIME type, tags
  • Sorting by date, size, relevance
  • Highlighted search results

Flow:

  1. Query is sent to Meilisearch
  2. Full-text search across indexed fields
  3. Filters are applied (dataset_id, status, etc.)
  4. Results are ranked and highlighted
  5. Metadata is returned (no S3 fetching)

Performance: Typically < 50ms for most queries

Use cases:

  • Keyword-based document discovery
  • Filtering by metadata (tags, type, dataset)
  • Search with typo tolerance
  • Faceted search interfaces
  • Finding documents by title, author, abstract

Request​

Responses​

Successful Response