How to run SQL queries in your Snowflake to do search for the taxonomy.? – RChilli HelpDesk

To run SQL queries in Snowflake for taxonomy-based searches related to RChilli's solution, follow the guidelines below based on your setup and goals. While RChilli itself does not directly offer SQL-based search over its taxonomy, organizations often import taxonomy or parsed data into a Snowflake data warehouse for internal analytics and querying. Here's how you can approach it:

Step-by-Step Guide: Running SQL Queries for Taxonomy in Snowflake

1. Import RChilli Parsed Output into Snowflake

RChilli outputs structured data (JSON/XML) after parsing resumes or job descriptions.
Use an ETL pipeline or Snowflake's COPY INTO command to load this data into Snowflake tables.

Example:

CREATE OR REPLACE TABLE parsed_resumes (
          candidate_id STRING,
          full_name STRING,
          skills ARRAY,
          job_profiles ARRAY,
          experience JSON,
          taxonomy JSON
  );

You can parse and load the "taxonomy" field from the JSON which typically includes skill aliases, categories, and mapped ontologies.

2. Query Taxonomy Data

Assuming your taxonomy-related data is stored in a column (e.g., taxonomy as JSON), you can extract and search using FLATTEN and LATERAL.

Example: Search for resumes with a specific skill alias:

SELECT
candidate_id,
full_name,
skill.value:FormattedName::STRING AS skill_name
FROM parsed_resumes,
LATERAL FLATTEN(input => parsed_resumes.taxonomy:Skills) AS skill
WHERE skill.value:FormattedName ILIKE '%JavaScript%';

This query:

Flattens the taxonomy.Skills array.
Filters results by skill name.

3. Search for Job Profiles by Ontology Mapping

Example:

SELECT
candidate_id,
job.value:FormattedName::STRING AS job_title,
job.value:Domain::STRING AS job_domain
FROM parsed_resumes,
LATERAL FLATTEN(input => parsed_resumes.taxonomy:JobProfiles) AS job
WHERE job.value:Domain = 'Information Technology';

Tips for Using Taxonomy in Snowflake

Preprocess JSON: Flatten JSON fields before querying large datasets.
Index key columns: Especially if you search frequently on skill names, domains, etc.
Normalize Skills: Consider storing aliases, synonyms, and categories in separate dimension tables for easier joins.
Use RChilli Taxonomy API for dynamic enrichment or additional mappings before loading data.

Related RChilli Resources

Need Help?

If you need guidance on setting up your data pipeline or specific SQL optimizations in Snowflake using RChilli taxonomy, feel free to reach out to support@rchilli.com