( AI, Data, Agentic AI, Cortex Code, NiFi, Iceberg, Polaris, Streamlit, Python, Java, SQL, MCP, LLM, RAG, Cortex AI, AISQL, Search, Unstructured Data )
Tim's Corner: May the Fourth Be With You
Welcome to issue 240 - and May the Fourth Be With You! Like the Rebel Alliance, we're building an open, interconnected data galaxy where information flows freely across systems.
This week the Force is strong with Snowflake: AI_PARSE_DOCUMENT leveled up with GA image extraction and 2,000-page support (that's a lot of Jedi archives),
Interactive Tables with CDC and streaming ingestion entered limited access (real-time data at lightspeed),
Data Security in Trust Center went GA (protecting the galaxy from the dark side of data breaches), and
SCIM authentication expanded with PATs and External OAuth.
The engineering team published a stellar blog on mining query history to build better semantic models for AI - turning your query logs into a knowledge base that makes AI assistants actually useful. On the community front, Prompt Forge landed for multi-model evaluation,
the Cortex Code Agent SDK got a managed services application showcase, and IBM released Granite 4.1 on Ollama.
I'm counting down to Snowflake Summit 2026 in San Francisco June 2-5 where I'll be speaking on NiFi and Agents. The countdown to Summit has begun - don't let this be a trap, register now! Let's dig in.
Tim's Content
Snowflake Summit 2026 - Tim is Speaking!
I'll be speaking at Snowflake Summit 2026 on NiFi and Agents. The Summit is less than a month away - this is your chance to join 10,000+ data professionals in San Francisco for the biggest data and AI event of the year. Check the session catalog:
New SNACK-AI Repositories
| Repository | Description |
|---|---|
| SNACKAI-CoCo-PremiumViews | Premium views and analytics built with Cortex Code |
| SLACKAI-CoCo-JetsonEdge | Edge AI on NVIDIA Jetson with Cortex Code and SNACK-AI |
| SNACKAI-CoCo-SnowflakeSummit2026Agenda | Summit 2026 session agenda explorer built with Cortex Code |
| SNACKAI-CoCo-CortexAITraining | Cortex AI training materials and examples built with SNACK-AI |
| SNACKAI-CoCo-AIGovernance | AI governance patterns built with Cortex Code and SNACK-AI |
| SNACKAI-CoCo-FindDataNearby | Location-aware data discovery built with Cortex Code |
| SNACKAI-CoCo-SpringSnowGhost | Spring ghost detection with Cortex Code and SNACK-AI |
| SNACKAI-CoCo-AIQueries | AI query builder with Cortex Code |
| SNACKAI-CoCo-PolyMarket | Prediction market analytics with Cortex Code |
| SNACKAI-CoCo-VPSCalculator | VPS cost calculator with Cortex Code |
Upcoming Events: Catch Tim Live!
- May 6: Philly Hands-On Lab: Automate Data Workflows with Cortex Code and Evolv - Philadelphia, PA
- May 13: Snowflake NoCode Live (SNL) - Virtual Hands-On Lab
- May 14: From Raw Data to a Conversational AI Agent with Snowflake in 90 Minutes - Virtual
- May 20: Budapest Data and AI - Budapest, Hungary - Tim Virtual speaking
- May 21: Data for Breakfast Encore - Virtual
- May 27: From Raw to Reliable: Build AI-Powered Data Quality Pipelines with Cortex Code - Virtual Hands-On Lab
- May 28: Build Your First Agentic ML Pipeline with Natural Language - Virtual Hands-On Lab
- June 2-5: Snowflake Summit 2026 - San Francisco, CA - Join me, I will be speaking on NiFi and Agents.
Snowflake Dev Day (In-Person): https://www.snowflake.com/en/dev-day/in-person/
All Snowflake Events: https://www.snowflake.com/about/events/
Past Events (Catch the Replays):
Snowflake Summit 2026 - San Francisco, June 2-5
The biggest event of the year is almost here! Snowflake Summit 2026 returns to San Francisco June 2-5. Snowflake announced that Anthropic's Co-Founder and President Daniela Amodei will join the keynote, along with 500+ sessions, 200+ on-site partners, and new AI-powered experiences. The Force will definitely be with you at Summit.
Press Release: Snowflake Makes AI Real at Summit 26
Agentic Enterprise Announcement: Snowflake Expands Intelligence and Cortex Code
Register now: https://www.snowflake.com/summit/
Cortex Code CLI Developer Trial: https://signup.snowflake.com/
Featured: Mining Query History to Build Better Semantic Models for AI
Giving an AI analyst raw database schemas and sample queries works great for demos but falls apart in production. Snowflake's engineering team published a deep dive on mining query history to build better semantic models - turning your existing query logs into a governed knowledge base that makes AI data assistants actually useful. The approach automatically extracts join patterns, filter logic, calculated metrics, and business terminology from how your team actually queries data.
Read more: Mining Query History to Build Better Semantic Models for AI
Featured: Interactive Tables - Real-Time Data at Lightspeed
Two new Limited Access features bring real-time interactive workloads to Snowflake:
Interactive CDC (Change Data Capture)
Use change data capture with interactive tables for real-time updates with much less ingestion cost compared to standard MERGE INTO patterns. Ideal for lower-volume, mutable data streams.
Documentation: https://docs.snowflake.com/en/LIMITEDACCESS/interactive-cdc
Interactive Streaming Ingestion
Stream data into interactive tables for real-time analytics. For mutable, deduplicated data (inserts, updates, deletes from source systems), combine with CDC for a complete real-time pipeline.
Documentation: https://docs.snowflake.com/en/LIMITEDACCESS/interactive-streaming
Cortex AI & Agents
AI_PARSE_DOCUMENT Image Extraction (GA - Apr 30)
AI_PARSE_DOCUMENT image extraction using LAYOUT mode is now GA. Extract images alongside text and layout to build multimodal document processing workflows. Unlock visual data from documents for richer AI applications.
AI_PARSE_DOCUMENT 2,000 Page Support (Apr 30)
AI_PARSE_DOCUMENT now processes documents up to 2,000 pages in both LAYOUT and OCR modes. Handle large healthcare records, insurance claims, regulatory filings, and technical manuals in a single request.
Read more: https://docs.snowflake.com/en/release-notes/2026/other/2026-04-30-ai-parse-document-2000-pages
Cortex Code: One Governed Agent for Your Entire Data Stack
Cortex Code expands into a governed agent across your entire data stack - warehouse, transformation, and orchestration. Available as Desktop, CLI, and in Snowsight.
Read more: Cortex Code Governed Agent Data Stack
Prompt Forge: Multi-Model Prompt Evaluation with Snowflake Cortex
Prompt Forge enables multi-model prompt evaluation directly in Snowflake Cortex - test prompts across different models, compare outputs, and optimize for quality and cost.
Resources:
Agentic Managed Services with Cortex Code Agent SDK
Build agentic managed services applications using the Cortex Code Agent SDK - demonstrates how to rebuild managed services workflows with AI agents.
Resources:
Awesome Pipeline Auditor with Cortex Agents SDK
A comprehensive pipeline auditor built with the Cortex Agents SDK for monitoring and validating data pipeline health.
GitHub: https://github.com/iamontheinet/awesome-pipeline-auditor-with-cortex-agents-sdk
Snowflake AI Kit: Cortex Code Plugin for Claude Code
Snowflake AI Kit now includes a Cortex Code plugin for Claude Code, bridging Snowflake's data context into Anthropic's coding assistant.
GitHub: https://github.com/Snowflake-Labs/snowflake-ai-kit#cortex-code-plugin-for-claude-code
TruLens 2.8.0 Released
TruLens 2.8.0 - the open source framework for evaluating, tracking, and optimizing AI (powers Snowflake's Agent GPA framework).
Release: https://github.com/truera/trulens/releases/tag/trulens-2.8.0
Cortex Code CLI & Agent SDK
Snowflake Platform Updates
Data Security in Trust Center (GA - Apr 24)
Data Security in the Trust Center is now GA. Automatic sensitive data classification across databases identifies PII, PCI, and regulated data categories - protecting the galaxy from the dark side of data breaches.
Read more: https://docs.snowflake.com/en/release-notes/2026/other/2026-04-24-data-security-trust-center-ga
SCIM Authentication: PATs and External OAuth (GA - May 1)
New authentication methods for SCIM requests are GA. Authenticate with Programmatic Access Tokens (PATs) and External OAuth tokens in addition to existing SCIM access tokens.
Read more: https://docs.snowflake.com/en/release-notes/2026/other/2026-05-01-scim-authentication
Outbound PrivateLink Endpoints Organization Usage View (May 1)
New OUTBOUND_PRIVATELINK_ENDPOINTS view in ORGANIZATION_USAGE schema for monitoring outbound private connectivity endpoints across all accounts in your organization.
Tags & Policies in Snowsight (Preview - Apr 30)
Create tags directly in Snowsight and view expanded tagging coverage across supported object types in the Tags & Policies Dashboard.
Read more: https://docs.snowflake.com/en/release-notes/2026/other/2026-04-30-snowsight-tags-policies-preview
Data Clean Rooms Updates (Apr 30)
Data Clean Rooms API Version 14.7 with creation timestamps for registry objects and Data Clean Rooms UI improvements.
Read more: https://docs.snowflake.com/en/release-notes/2026/other/2026-04-30-dcr
Datometry for Snowflake: Accelerate Teradata Migration
Datometry for Snowflake removes the difficult tradeoff for enterprises on Teradata - no more choosing between expensive renewals or high-risk multi-year rewrites. Accelerate your exit to the AI Data Cloud.
Read more: Announcing Datometry for Snowflake
Snowpark Connect for Apache Spark
Snowpark Connect for Apache Spark enables Spark workloads to run directly on Snowflake compute, eliminating data movement.
Product Page: https://www.snowflake.com/en/product/features/snowpark-connect-apache-spark/
Booking.com Customer Story
See how Booking.com leverages Snowflake for data and AI at scale.
Video: https://www.snowflake.com/en/customers/all-customers/video/booking-com/?clear-cache=59tbik
OpenFlow
OpenFlow Connector for Confluence Data Center
Openflow Connector for Confluence Data Center ingests wiki and knowledge base content into Snowflake for search and analytics.
Documentation: https://docs.snowflake.com/en/LIMITEDACCESS/openflow/connectors/confluence-data-center/about
EvanFlow
Community-built OpenFlow connector patterns and utilities.
GitHub: https://github.com/evanklem/evanflow
Dynamic Tables & dbt
Migrating a dbt Core Project to Snowflake
Guide on migrating a dbt Core project to Snowflake - from setup through deployment patterns.
Read more: https://www.linkedin.com/pulse/migrating-dbt-core-project-snowflake-john-nielsen-xsxfc/
Dynamic Tables + dbt: A Powerful Combination
How Dynamic Tables and dbt work together for a powerful data transformation pipeline.
Read more: https://medium.com/snowflake/dynamic-tables-dbt-a-powerful-combination-f550ebc23d60
AI Models & Research
IBM Granite 4.1
IBM Granite 4.1 is now available on Ollama - an open, enterprise-focused model family with strong tool-use and reasoning capabilities.
Resources:
Moondream v2
Moondream v2 - a lightweight vision-language model for edge and embedded use cases, available on Ollama.
Ollama: https://ollama.com/library/moondream:v2
Frontier Model Lower Costs
Analysis of how frontier model costs continue to decline, making AI more accessible.
Read more: https://www.mendral.com/blog/frontier-model-lower-costs
Microsoft Agent Framework
Microsoft's Agent Framework - an open source framework for building, managing, and orchestrating AI agents.
GitHub: https://github.com/microsoft/agent-framework
Generative AI for Science
Research repository on applying Generative AI to scientific research and discovery.
GitHub: https://github.com/jpliu168/Generative_AI_For_Science/
From the Community: Medium & Blogs
| Article | Description |
|---|---|
| Snowflake Intelligence: The Fifth Person | How Snowflake Intelligence solves the business user data access problem |
| Production-Grade AI Agent Inside Snowflake | Every line of building a production AI agent in Snowflake |
| Enterprise-Grade skill.md for Cortex Code (Part 3) | Scaling and polishing enterprise skill definitions |
| Enterprise-Grade skill.md for Cortex Code (Part 2) | Building the workflow for enterprise skills |
| Enterprise-Grade skill.md for Cortex Code (Part 1) | Foundation patterns for production skill definitions |
| Prompt Forge: Multi-Model Evaluation | Test prompts across models with Snowflake Cortex |
| The End of the Billable Hour: Cortex Code Agent SDK | Rebuilding managed services with AI agents |
| Dynamic Tables + dbt: A Powerful Combination | How Dynamic Tables and dbt work together |
| Snowflake Trust Center Detections | Real-time threat monitoring with Trust Center |
| Snowflake Database Roles Demystified | Understanding who sees what and why it matters |
| From All-or-Nothing to Graceful Degradation: DML Error Logging | Graceful degradation patterns for DML error handling |
| Hybrid Data Mesh with Snowflake and Databricks | Building a hybrid data mesh across platforms |
GitHub: Snowflake Labs & Community
| Repository | Description |
|---|---|
| snowflake-ai-kit | Snowflake AI Kit with Cortex Code plugin for Claude Code |
| prompt_forge | Multi-model prompt evaluation with Snowflake Cortex |
| Agentic-Managed-Services | Managed services application with Cortex Code Agent SDK |
| awesome-pipeline-auditor | Pipeline auditor with Cortex Agents SDK |
| Snow-Insurance-Demo | Snowflake insurance industry demo application |
| VibeCodingExamples | Snowflake vs Databricks Spark TPC-H SF100 ETL comparison |
| trulens v2.8.0 | TruLens 2.8.0 - AI evaluation framework powering Agent GPA |
| evanflow | Community OpenFlow connector patterns |
| SNACKAI-CoCo-PremiumViews | Premium views with Cortex Code |
| SLACKAI-CoCo-JetsonEdge | Edge AI on NVIDIA Jetson |
Developer Tools & Open Source
| Tool | Description |
|---|---|
| datacenter.fm | Interactive map and data on global data center infrastructure |
| datacenter.fm (GitHub) | Open source data center mapping project |
| cua | Computer-use agent framework for automating desktop tasks |
| ministack | Minimal infrastructure stack for local development |
| named-pipes | Named pipes utilities for inter-process communication |
| open-pdf-studio | Open source PDF editing and processing studio |
| winpodx | Windows container pod management |
| utilyze | System utilization monitoring and analytics |
| Systalyze Utilyze | Cloud resource utilization platform |
| powermonitor | Power consumption monitoring tool |
| maxheadbox | Max Headroom-inspired digital art project |
| l123 | Lightweight data processing utility |
| Meshtastic Apple | Meshtastic mesh networking app for Apple devices |
| warmburnout.com | Platform comparison and burnout prevention resources |
| Doom in ChatGPT and Claude | Running Doom inside AI chat interfaces |
| Terminal GPU Battery (Ghostty/iTerm2) | GPU and battery optimization for Mac terminal emulators |
Video Content
| Video | Description |
|---|---|
| Cortex Code Agent SDK Demo | Demo of Cortex Code Agent SDK capabilities |
| Snowflake AI & ML Playlist | Snowflake AI and ML video series |
SNACK-AI: The 2026 Pattern
As we look forward in 2026, SNACK-AI is the architectural pattern for powering AI applications at any scale - from mobile phones and robots to massive enterprise clusters.
SNACK-AI integrates:
- Snowflake Platform
- NiFi (Apache)
- Apache Iceberg
- Cortex AI (Agents/Search/MCP)
- Kafka (Apache Kafka)
Deep Dive into SNACK-AI:
- Read: SNACK-AI: The 2026 Pattern (Medium)
- Code: Official SNACK-AI GitHub Repository
- Weekly Context: SNACK-AI Overview
Thanks
https://github.com/timothyspann
May the Fourth Be With You! Generated with Cortex Code for All Data and AI Weekly - May 4, 2026
Top comments (0)