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 )

NiFi + AI + AI Data Cloud + Iceberg.
https://www.reddit.com/r/DataEngineeringForAI/hot/
https://github.com/tspannhw/TrafficAI/tree/main/Agents
https://github.com/tspannhw/conferences

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.
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.
- 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
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:
- 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.
- 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.
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)
- Kafca (Apache Kafka)
- Read: SNACK-AI: The 2026 Pattern (Medium)
- Code: Official SNACK-AI GitHub Repository
- Weekly Context: SNACK-AI Overview
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.
The journey doesn't stop here. We are already building the foundation for a single source of truth for activity data.
- Connect: Building the Single Source of Truth (LinkedIn)
- Watch: AI & Data Predictions for 2026 (Webinar)
See you in 2026!

https://github.com/timothyspann
© 2020-2025 Tim Spann https://www.youtube.com/@FLaNK-Stack


