Helpful Snowflake Tutorials
September 9, 2026 · PowderInsights Team
An end-to-end in-platform pipeline for PDFs and scans: directory-table stages, AI_PARSE_DOCUMENT modes, training a Document AI extraction model, confidence thresholds and review queues, incremental streams and tasks, and a Cortex Search index over the text.
Read moreSeptember 9, 2026 · PowderInsights Team
Stop deploying from worksheets. A working Snowflake DevOps setup: Git repositories inside Snowflake, idempotent CREATE OR ALTER object files, zero-copy clone PR environments, the Snowflake CLI, a GitHub Actions workflow, and honest rollback options.
Read moreSeptember 9, 2026 · PowderInsights Team
Passwords alone no longer work in Snowflake. A practical hardening tutorial: classifying users with TYPE, authentication and password policies, key-pair rotation, scoped programmatic access tokens, workload identity federation for AWS and GitHub Actions, least-privilege service roles, and the audit queries that prove it.
Read moreSeptember 9, 2026 · PowderInsights Team
A hands-on guide to moving PySpark and pandas ETL onto Snowpark for Python: DataFrame API mappings, Snowpark pandas (Modin), UDFs and UDTFs, stored procedures, warehouse sizing, and the migration patterns that actually hold up in production.
Read moreSeptember 9, 2026 · PowderInsights Team
Snowflake now has two ways to run row-oriented, low-latency workloads: Hybrid Tables with real primary keys and row locking, and managed Snowflake Postgres. Working SQL, the limits of each, and a decision rule for picking between them and standard tables.
Read more