All Data and AI Weekly #222-29 Dec 2025

#llm #mcp #genai #ai

All Data and AI Weekly #222-29 Dec 2025

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

🚀 All Data and AI Weekly: Best of 2025 & The Road to 2026

Welcome to this week's edition of the newsletter. As we wrap up an incredible year of innovation, we are looking back at the definitive "Best of 2025" and casting our gaze forward toward the architectural patterns that will define 2026.

🏆 The Best of 2025: A Year in Review

2025 was the year real-time AI moved from experimental to essential. We’ve consolidated the most impactful resources, playbooks, and summaries to help you audit your current stack and prepare for the next wave.

Key Highlights:

The Real-Time AI Playbook: Your strategic guide to deploying low-latency intelligence.
2025 Summary & Unwrapped: A deep dive into the milestones of the FLiPStack ecosystem.
Lakehouse Pipeline Repository: The blueprint for modern, scalable data engineering.

Explore the full archive here: Best of 2025 GitHub

❄️ Snowflake: The "New Hotness"

The Snowflake ecosystem has evolved rapidly this year, shifting from a data warehouse to a full-scale AI and application platform. Here is the latest tech you should be tracking:

AI & Intelligence

Snowflake Intelligence: Enterprise-grade AI at your fingertips.
Cortex Agents & Knowledge Extensions: Building agentic workflows with ease.
Model Context Protocol (MCP) Server: Standardizing how agents interact with your data.
Cortex AISQL: Structuring unstructured data through natural language.

Engineering & Infrastructure

SnowConvert AI: Accelerated cloud migrations.
Dynamic Tables & Snowpipe Streaming: High-performance, declarative data pipelines.
Pg_lake: Seamless Postgres-to-Lakehouse integration.
Openflow: The new standard for orchestration and workflow management.

🍭 Looking Ahead to 2026: SNACK-AI

As we look toward 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:

Snowflake Platform
NiFi (Apache)
Aceberg (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

🛠️ Late 2025 Tooling & Updates

Don't miss these critical updates and community projects that landed in the final months of the year:

Cost Efficiency: SiS Container Runtime — Run Streamlit apps at a fraction of the cost.
Data Patterns: Retiring the Deferred Table Framework — A new pattern for late/out-of-order events.
Security: Prompt Injection Detection with Snowflake and Hugging Face.
Python/Data: Iceframe — Polars-like API for Iceberg.
Open Source: Uptime-Kita and Tunnelto.