Semantic Search

Point-in-time and utility-filtered passage retrieval over Ultralayer's financial corpus (1-8 queries).

post/v0/search/semantic_search

Request

queriesstring[] required

max_tokensinteger

Snippet token budget

start_timestampstring date-time nullable

System timestamp; Inclusive ISO 8601

end_timestampstring date-time nullable

System timestamp; Inclusive ISO 8601; null=realtime

min_cosine_similaritynumber nullable

entity_typesstring[] nullable

Match any.

min_utility_scorenumber nullable

0-.2 noise; .2-.4 low; .4-.6 mod; .6-.8 high; .8-1 critical

min_sentiment_scorenumber nullable

-1..1; +-.2 neutral; +-.2-.6 mod; +-.6+ strong; mutually exclusive with max_sentiment_score

max_sentiment_scorenumber nullable

detail'full' | 'standard' | 'essential'

essential:minimal; standard:citable; full:all.

{
  "queries": [
    "Apple iPhone demand outlook",
    "Apple services revenue growth",
    "Apple CEO guidance"
  ]
}

Semantic search results at detail=full.

timestampstring date-time required

System imestamp of the document.

document_idinteger required

Identifier of the indexed document.

titlestring nullable required

Title of the source document, if available.

snippetsstring[] required

Passage texts from this document, ordered as they appear in the document.

snippet_cosine_similaritiesnumber[] required

Cosine similarity of each snippet to the query. Same length and order as snippets.

o200k_base_n_tokensinteger required

o200k_base token count of the returned snippets.

canonical_entity_namesstring[] required

Canonical entity names across the returned snippets.

utility_scorenumber required

Token-weighted utility score of the returned snippets, between 0.0 and 1.0.

sentiment_scorenumber required

Token-weighted sentiment score of the returned snippets, between -1.0 and 1.0.

No recorded changes to this operation across all 1 revision of this API.