Pipelinepipeline · returns table

FILTER

Filter query results using LLM-based semantic matching

Table-in, table-out — composes downstream of SELECTs.

pipelinellmpipeline-composabletext

Syntax

THEN FILTER {{ prompt }}

THEN FILTER({{ prompt }})

Arguments

name	type	description
prompt	VARCHAR	—
_table	TABLE	—

About

PIPELINE cascade for LLM-based filtering of query results. Used with THEN FILTER syntax for semantic filtering: SELECT * FROM products THEN FILTER('only eco-friendly items') Uses LLM to evaluate each row against the filter criteria and returns only matching rows with the SAME SCHEMA as input.

Examples

Keeps only rows that clearly match the semantic filter

SELECT
  *
FROM
  (
    VALUES
      ('Reusable bamboo cup'),
      ('Single-use plastic straw'),
      ('Recyclable paper bag')
  ) AS t (description) THEN FILTER (
    'keep rows that explicitly describe eco-friendly, reusable, or recyclable items'
  ) THEN PYTHON ('result = pd.DataFrame({"row_count":[len(df)]})')

Nearby rabbit holes

same domain

pipeline

ADD_STYLES

Apply theme styling to a chart specification

pipelinellmpipeline-composabletext

pipeline

ANALYZE

Analyze query results with LLM based on a prompt

pipelinellmpipeline-composabletext

pipeline

DEDUPE

Remove duplicate rows

pipelinellmpipeline-composabletext

pipeline

ENRICH

Add LLM-computed columns to query results

pipelinellmpipeline-composabletext

pipeline

GROUP

Group by column and aggregate another

pipelinellmpipeline-composabletext

pipeline

INVESTIGATE

Investigative analysis - explores related data to answer questions

pipelinellmpipeline-composabletext

Climb back to The Looking Glass

surface↓pipeline↓filter

Pipelinepipeline · returns table

FILTER

Filter query results using LLM-based semantic matching

Table-in, table-out — composes downstream of SELECTs.

pipelinellmpipeline-composabletext

Syntax

THEN FILTER {{ prompt }}

THEN FILTER({{ prompt }})

Arguments

name	type	description
prompt	VARCHAR	—
_table	TABLE	—

About

Examples

Keeps only rows that clearly match the semantic filter

SELECT
  *
FROM
  (
    VALUES
      ('Reusable bamboo cup'),
      ('Single-use plastic straw'),
      ('Recyclable paper bag')
  ) AS t (description) THEN FILTER (
    'keep rows that explicitly describe eco-friendly, reusable, or recyclable items'
  ) THEN PYTHON ('result = pd.DataFrame({"row_count":[len(df)]})')