Timothy Spann

Posted on Mar 2

📰 All Data and AI Weekly #231-02March2026

#mcp #ai #genai #agenticai

📰 All Data and AI Weekly #231-02 March 2026

( AI, Data, NiFi, Iceberg, Polaris, Streamlit, Flink, Kafka, Python, Java, SQL, MCP, LLM, RAG, Cortex AI, AISQL, Search, Unstructured Data )

Blizzard!!! So much snow here.

🌟 Tim's Corner: SNACKAI Innovations

Recent dashboards and pipelines from Tim Spann.

SNACKAI CoCo Dashboards: Visualizing real-time AI and connectivity data.
SNACKAI CoCo DBT-MTA Pipeline: Advanced data modeling for transit analytics.
SNACKAI CoCo NYC OPSCenter: Operations center monitoring for municipal data.

🗓️ Upcoming Events: Catch Tim Live!

For in-person I have a new sticker!

March 3: NYC Interactive
March 4: AI and Data Sharing
March 6: LLM Days NYC - NYC Tim Speaking
March 11: MCP Conference NYC – NYC Tim Speaking
March 12: Data for Breakfast NYC – NYC Hands on Labs
March 18: Data for Breakfast Princeton - Princeton
March 19: Warsaw IT Days - Virtual
March 24: NYC Snowflake Users Group - NYC Hands on Workshop
April 15: Nodes AI Conference - Virtual
May 20: Budapest Data and AI – Budapest, Hungary - Tim Virtual speaking

https://www.snowflake.com/about/events/

Coupon for MCP Conference

https://luma.com/qi5g9hke?coupon=SPEAKER%20SOCIAL%2050%25

“From TrafficAI to GhostBreakers: Building Stateful AI Agents with Cortex, OpenFlow, and Snowflake Postgres”

Next week, I’ll be presenting at LLMday New York Q1 on 6 March.
I’ll share how to build stateful AI agents inside Snowflake that operate on live data, coordinate across workflows, and hold context across complex tasks.

Secure your spot at 50% off with my code LASTMINUTE

The 17th edition of Warsaw IT Days is taking place online on March 19th and at the PGE National Stadium in Warsaw on March 20th, and I have the pleasure of participating as a speaker. In addition to my presentation, you can also enjoy over 300 lectures across 25 different thematic tracks.

Join and hear about the latest trends and solutions in IT and data science, take advantage of the opportunity to deepen your technical knowledge, and exchange experiences with a community of over 10,000 industry specialists and managers.

More information and registration at: https://warszawskiedniinformatyki.pl/en/

https://hs.streamnative.io/webinar-introducing-the-streamnative-remote-mcp-server?ut

❄️ The Snowflake Ecosystem

Snowflake is moving fast with Cortex, OpenFlow, and Apache Iceberg.

Cortex & AI Capabilities

Cortex Code CLI Expansion: Now supporting more workflows for developers.
Advanced Cortex Skills: Deep dive into Code CLI extensibility.
Cortex Agent Evaluations: Measuring the performance of your AI agents.
MLflow-like LLM Evaluation: Native evaluation tools within the platform.
Snowflake Copilot Guide: Getting started with the built-in AI assistant.

Data Engineering & Connectivity

Replacing Debezium with OpenFlow: A new paradigm for data movement.
OpenFlow Explained: A human body analogy for data flow.
Streamlining Postgres to Snowflake: Simplifying the migration and sync process.
Iceberg Tables & Hierarchical Paths: Enhancing partitioning in the open lakehouse.
Row Timestamps: Better tracking of data changes.

Semantic Layers & Analytics

Semantic View Autopilot: Automating the creation of semantic layers.
Best Practices for Cortex Analyst: Optimizing for natural language queries.
End-to-End Analytics with Tableau: Leveraging semantic views for BI.

🔌 The Rise of MCP (Model Context Protocol)

The new standard for connecting AI models to data sources.

Official Google MCP Repository: The core protocol documentation and SDKs.
Awesome MCP Servers: A curated list of servers to extend your AI.
Postgres MCP: Connecting LLMs directly to your Postgres databases.
MS Copilot Studio & Snowflake MCP: Bridging Microsoft and Snowflake via MCP.
MCP-Use: Practical implementations of the protocol.

🛠️ Open Source & Developer Tools

Essential repos and tools for the modern data stack.

Pandas-AI: Talk to your dataframes in natural language.
Kreuzberg: A modern library for text extraction.
MegaParse: The ultimate parser for LLMs.
Polaris Local Forge: Local development for Apache Polaris.
SQL Studio: A SQL database explorer that runs in the browser.
PG Dog: Postgres replication and load balancing.
Traccar: Modern GPS tracking system.
FingerprintJS: Browser fingerprinting for fraud detection.

🧠 AI Research & Multi-Agent Systems

Where the industry is heading next.

Spotify’s Multi-Agent Architecture: How Spotify uses agents for smarter advertising.
Graph Agents for Snowflake: Integrating knowledge graphs with AI agents.
Raptor: Recursive Abstractive Processing for Tree-Organized Retrieval.
Ironclaw: High-performance AI reasoning frameworks.

📊 Deep Dive: Data Engineering & Storage

Formats, parsers, and performance.

Geospatial Types in Parquet: Native support for spatial data is finally here.
Hardwood Parser: A high-performance Java parser for Parquet.
Apache DataSketches: A library for approximate queries on big data.
Postgres JSONB & TOAST: Understanding storage internals.
DuckDB & Apache Arrow: Speeding up PySpark workloads.

📺 Community & Events

Snowflake Connect AI EMEA: Check out the upcoming agenda.
Observe Podcast: How Observe uses Iceberg and Snowflake.
State of Data Engineering: A comprehensive repo on the current industry landscape.
AI Improving AI Video: Insightful discussion on self-improving systems.

Quick Links

📜 Historical Throwback

Looking back to see how far we've come.

Deep Learning on HDP (2018): A look back at running neural networks on Hortonworks Data Platform in Prague.

🍭 2026: SNACK-AI

As we look forward in 2026, we are introducing SNACK-AI. This architectural pattern is designed to power AI applications at any scale—from mobile phones and robots to massive enterprise clusters.

SNACK-AI integrates: