SQL Keywords

AI_FILTER

Feb 23, 2026

min read

AI_FILTER

Overview

Returns True or False for a given text or image input based on a natural language condition. Use in SELECT, WHERE, or JOIN clauses to filter results.

Syntax

AI_FILTER(
  condition,
  input
)

AI_FILTER(
  condition,
  input
)

AI_FILTER(
  condition,
  input
)

Parameters

condition (VARCHAR): Natural language description of the filter condition
input (VARCHAR or FILE): Text string or file reference to evaluate

Use Cases

Content moderation and filtering
Data quality checks
Conditional data processing
Smart data filtering
Policy compliance checking
Image content filtering

Code Examples

Example 1: Simple Text Filtering

SELECT 
    country,
    AI_FILTER('Is this country in Asia?', country) AS is_in_asia
FROM

SELECT 
    country,
    AI_FILTER('Is this country in Asia?', country) AS is_in_asia
FROM

SELECT 
    country,
    AI_FILTER('Is this country in Asia?', country) AS is_in_asia
FROM

Output:

country   | is_in_asia
----------|------------
Japan     | true
Germany   | false
Thailand  | true
Canada    | false

country   | is_in_asia
----------|------------
Japan     | true
Germany   | false
Thailand  | true
Canada    | false

country   | is_in_asia
----------|------------
Japan     | true
Germany   | false
Thailand  | true
Canada    | false

Example 2: Filter in WHERE Clause

SELECT 
    product_name,
    description,
    price
FROM products
WHERE AI_FILTER(
    'Does this product description mention eco-friendly or sustainable materials?',
    description
) = true

SELECT 
    product_name,
    description,
    price
FROM products
WHERE AI_FILTER(
    'Does this product description mention eco-friendly or sustainable materials?',
    description
) = true

SELECT 
    product_name,
    description,
    price
FROM products
WHERE AI_FILTER(
    'Does this product description mention eco-friendly or sustainable materials?',
    description
) = true

Example 3: Customer Review Filtering

SELECT 
    review_id,
    review_text,
    AI_FILTER(
        'Does this review contain complaints about shipping or delivery?',
        review_text
    ) AS has_shipping_complaint
FROM customer_reviews
WHERE created_date >= CURRENT_DATE - 30

SELECT 
    review_id,
    review_text,
    AI_FILTER(
        'Does this review contain complaints about shipping or delivery?',
        review_text
    ) AS has_shipping_complaint
FROM customer_reviews
WHERE created_date >= CURRENT_DATE - 30

SELECT 
    review_id,
    review_text,
    AI_FILTER(
        'Does this review contain complaints about shipping or delivery?',
        review_text
    ) AS has_shipping_complaint
FROM customer_reviews
WHERE created_date >= CURRENT_DATE - 30

Example 4: Multi-Condition Filtering

SELECT 
    email_id,
    subject,
    body
FROM emails
WHERE AI_FILTER('Is this email urgent or high priority?', subject || ' ' || body) = true
  AND AI_FILTER('Does this email require immediate action?', body) = true

SELECT 
    email_id,
    subject,
    body
FROM emails
WHERE AI_FILTER('Is this email urgent or high priority?', subject || ' ' || body) = true
  AND AI_FILTER('Does this email require immediate action?', body) = true

SELECT 
    email_id,
    subject,
    body
FROM emails
WHERE AI_FILTER('Is this email urgent or high priority?', subject || ' ' || body) = true
  AND AI_FILTER('Does this email require immediate action?', body) = true

Example 5: Image Content Filtering

SELECT 
    file_name,
    AI_FILTER(
        'Does this image contain people?',
        TO_FILE('@images/' || file_name)
    ) AS contains_people
FROM DIRECTORY('@images')
WHERE file_name LIKE '%.jpg'

SELECT 
    file_name,
    AI_FILTER(
        'Does this image contain people?',
        TO_FILE('@images/' || file_name)
    ) AS contains_people
FROM DIRECTORY('@images')
WHERE file_name LIKE '%.jpg'

SELECT 
    file_name,
    AI_FILTER(
        'Does this image contain people?',
        TO_FILE('@images/' || file_name)
    ) AS contains_people
FROM DIRECTORY('@images')
WHERE file_name LIKE '%.jpg'

Example 6: Dynamic Filtering with PROMPT

SELECT 
    product_category,
    AI_FILTER(
        PROMPT('Is {0} a category related to technology?', product_category),
        product_category
    ) AS is_tech_related
FROM

SELECT 
    product_category,
    AI_FILTER(
        PROMPT('Is {0} a category related to technology?', product_category),
        product_category
    ) AS is_tech_related
FROM

SELECT 
    product_category,
    AI_FILTER(
        PROMPT('Is {0} a category related to technology?', product_category),
        product_category
    ) AS is_tech_related
FROM

Data Output Examples

Policy Compliance

Input: "This document contains confidential financial information"
Condition: "Does this text indicate confidential or sensitive information?"
Output: true

Input: "This document contains confidential financial information"
Condition: "Does this text indicate confidential or sensitive information?"
Output: true

Input: "This document contains confidential financial information"
Condition: "Does this text indicate confidential or sensitive information?"
Output: true

Quality Checks

Input: "This product is made from 100% recycled materials"
Condition: "Does this mention environmental sustainability?"
Output: true

Input: "This product is made from 100% recycled materials"
Condition: "Does this mention environmental sustainability?"
Output: true

Input: "This product is made from 100% recycled materials"
Condition: "Does this mention environmental sustainability?"
Output: true

Model Information

Model Used: Snowflake managed model
Context Window: 128,000 tokens
Output: Boolean (true/false)

Limitations & Considerations

Performance

Best for batch processing, not real-time filtering
Use MEDIUM warehouse or smaller
Consider caching results for repeated filters

Cost

Billed per input processed
Condition text counts as input tokens for each row
Use concise, clear conditions

Accuracy

Works best with specific, unambiguous conditions
May struggle with highly subjective criteria
Test with sample data before large-scale use

Regional Availability

AWS US West/East: ✓
Azure East US: ✓
EU regions: ✓
Cross-region inference: ✓

Best Practices

1. Write Clear Conditions

-- Good: Specific and unambiguous
AI_FILTER('Does this text contain profanity or offensive language?', text)

-- Less effective: Vague
AI_FILTER('Is this bad?', text)

-- Good: Specific and unambiguous
AI_FILTER('Does this text contain profanity or offensive language?', text)

-- Less effective: Vague
AI_FILTER('Is this bad?', text)

-- Good: Specific and unambiguous
AI_FILTER('Does this text contain profanity or offensive language?', text)

-- Less effective: Vague
AI_FILTER('Is this bad?', text)

2. Use PROMPT for Dynamic Conditions

-- Build dynamic conditions
SELECT 
    product_name,
    AI_FILTER(
        PROMPT('Does {0} contain the keyword {1}?', product_name, 'organic'),
        product_name
    ) AS matches
FROM

-- Build dynamic conditions
SELECT 
    product_name,
    AI_FILTER(
        PROMPT('Does {0} contain the keyword {1}?', product_name, 'organic'),
        product_name
    ) AS matches
FROM

-- Build dynamic conditions
SELECT 
    product_name,
    AI_FILTER(
        PROMPT('Does {0} contain the keyword {1}?', product_name, 'organic'),
        product_name
    ) AS matches
FROM

3. Combine with Traditional Filters

-- Use AI_FILTER for complex logic, traditional WHERE for simple filters
SELECT *
FROM reviews
WHERE rating <= 2  -- Traditional filter
  AND AI_FILTER('Does this mention product defects or quality issues?', review_text) = true

-- Use AI_FILTER for complex logic, traditional WHERE for simple filters
SELECT *
FROM reviews
WHERE rating <= 2  -- Traditional filter
  AND AI_FILTER('Does this mention product defects or quality issues?', review_text) = true

-- Use AI_FILTER for complex logic, traditional WHERE for simple filters
SELECT *
FROM reviews
WHERE rating <= 2  -- Traditional filter
  AND AI_FILTER('Does this mention product defects or quality issues?', review_text) = true

4. Consider Materialized Views

-- For frequently used filters, materialize results
CREATE MATERIALIZED VIEW eco_products AS
SELECT 
    product_id,
    product_name,
    AI_FILTER('Is this product environmentally friendly?', description) AS is_eco_friendly
FROM

-- For frequently used filters, materialize results
CREATE MATERIALIZED VIEW eco_products AS
SELECT 
    product_id,
    product_name,
    AI_FILTER('Is this product environmentally friendly?', description) AS is_eco_friendly
FROM

-- For frequently used filters, materialize results
CREATE MATERIALIZED VIEW eco_products AS
SELECT 
    product_id,
    product_name,
    AI_FILTER('Is this product environmentally friendly?', description) AS is_eco_friendly
FROM

Common Use Cases

Content Moderation

SELECT comment_id, comment_text
FROM user_comments
WHERE AI_FILTER('Does this contain hate speech or harassment?', comment_text) = false

SELECT comment_id, comment_text
FROM user_comments
WHERE AI_FILTER('Does this contain hate speech or harassment?', comment_text) = false

SELECT comment_id, comment_text
FROM user_comments
WHERE AI_FILTER('Does this contain hate speech or harassment?', comment_text) = false

Compliance Checking

SELECT document_id
FROM legal_documents
WHERE AI_FILTER('Does this document comply with GDPR requirements?', content) = true

SELECT document_id
FROM legal_documents
WHERE AI_FILTER('Does this document comply with GDPR requirements?', content) = true

SELECT document_id
FROM legal_documents
WHERE AI_FILTER('Does this document comply with GDPR requirements?', content) = true

Smart Tagging

SELECT 
    article_id,
    title,
    AI_FILTER('Is this article about artificial intelligence?', title || ' ' || content) AS is_ai_topic
FROM

SELECT 
    article_id,
    title,
    AI_FILTER('Is this article about artificial intelligence?', title || ' ' || content) AS is_ai_topic
FROM

SELECT 
    article_id,
    title,
    AI_FILTER('Is this article about artificial intelligence?', title || ' ' || content) AS is_ai_topic
FROM

Related Functions

AI_CLASSIFY - For multi-category classification
AI_SENTIMENT - For sentiment analysis
PROMPT - For building dynamic conditions
AI_COMPLETE - For complex evaluations

Interested to Learn More?
Try Out the Free 14-Days Trial

Start free trial

Learn

Feb 18, 2026

BigQuery Global Queries: How to Run Cross-Region SQL in 2026

Analytics

Feb 18, 2026

Context Engineering and AI Quality for Data Teams

Product

Feb 18, 2026

Accelerate Analytics Development with Paradime and Tableau

Product

Feb 18, 2026

Accelerate Analytics Development with Paradime and Tableau

Experience Analytics for the AI-Era

Start your 14-day trial today - it's free and no credit card needed

Start for free

Experience Analytics for the AI-Era

Start your 14-day trial today - it's free and no credit card needed

Start for free

Experience Analytics for the AI-Era

Start your 14-day trial today - it's free and no credit card needed

Start for free

Platform

Resources

ADD-ONs

Industries

About

Legal

Made with ❤️ in San Francisco ・ London

*dbt® and dbt Core® are federally registered trademarks of dbt Labs, Inc. in the United States and various jurisdictions around the world. Paradime is not a partner of dbt Labs. All rights therein are reserved to dbt Labs. Paradime is not a product or service of or endorsed by dbt Labs, Inc.

Start for free

Platform

Resources