surface↓classification↓timeframe_llm
Classificationdimension · returns varchar

TIMEFRAME_LLM

LLM-backed timeframe extraction (escape hatch for TIMEFRAME)

Per-row classifier — stable across GROUP BY.

classificationllmllm-escape-hatchtext

Arguments

nametypedescription
textVARCHAR—
granularityVARCHAR—

About

LLM-backed alternative to TIMEFRAME for extracting when events in a text occurred or will occur. Reach for this when the base NLI operator can't handle subtle temporal cues — relative phrasing ("three weeks after the last incident"), narrative time-shifts, or event sequencing that needs reasoning rather than cue-matching. Useful on legal, medical, or historical text where the timeline is implicit in the prose. Shares the same `granularity` argument as TIMEFRAME (time_of_day, season, weather, duration). For routine timestamp extraction from structured or semi-structured text, TIMEFRAME on the NLI path is much cheaper.

Examples

LLM escape hatch extracts timeframe

SELECT
  timeframe_llm ('Expected to launch next quarter')

Nearby rabbit holes

same domain
dimension

AUDIENCE

Identify target audience via zero-shot NLI

classificationnlispecialist-zootext+1
dimension

AUDIENCE_LLM

LLM-backed audience identification (escape hatch for AUDIENCE)

classificationllmllm-escape-hatchtext
dimension

AUTHENTICITY

Assess authenticity of content via zero-shot NLI

classificationnlispecialist-zootext
dimension

AUTHENTICITY_LLM

LLM-backed authenticity assessment (escape hatch for AUTHENTICITY)

classificationllmllm-escape-hatchtext
dimension

BUCKET

Classify text into user-specified buckets via zero-shot NLI

classificationnlispecialist-zootext
dimension

BUCKET_LLM

LLM-backed bucketing (escape hatch for BUCKET)

classificationllmllm-escape-hatchtext
Climb back to The Looking Glass