Inmon's Data Warehouse Blues, Google to buy Anthropic? #116 w/e 2 Jan 2026
Join the 6,500-strong data herd getting all you need to know about Data for your Friday roundup
This was a quiet week as the New Year eases itself in, but there were rumors wandering around of Google setting up to purchase Anthropic.
They have invested more than previously thought, owning about 14% (source).
It is hard to tell if the Snowflake / Observe Inc, acquisition will have legs but it is remarkably quiet given the various Modern data stack acquisitions over the last year got so much attention. latest
And why is noone talking about AtScale?
Orchestra Product Updates
Are you tired of spinning up multiple Airflow clusters for different teams? Do you worry about how to keep governance boundaries clear and ensure noone can see things they can’t? Well worry no more!
You can now create multiple Workspaces/Accounts in Orchestra. This means that you can easily create different areas for different teams, without having to spin-up any extra infrastructure. This applies to environments as well, so creating a DEV, STAGE and PROD environment for multiple teams no longer means you need to create a ton of infrastructure.
December Feature Log
dbt Core jobs can assume IAM Roles (helpful if you are running dbt Loom in an enterprise setting so you can go and get artifacts from other repos via e.g. S3)
Claude Integration
Hybrid Compute for Azure using Azure Container Applications
Data Quality Tests for Postgres
Secret Data transformation feature
Orchestra supports Running Jobs in Dataproc
Orchestra supports sending alerts to PagerDuty
Orchestra supports parameterisation of Cloud Run
Orchestra supports parameterised triggering of Synapse Notebooks
Additional AI Agents in the Agents Tab (docs)
Parameters for Google Cloud Run Jobs (docs)
Google Cloud Run Functions (docs)
Improved Metadata Tracking - if you head to the Assets panel and set-up an Asset Run, Orchestra now fetches lineage and columnar metadata such as descriptions, types etc. Currently only supported for Snowflake.
Cancelling Pipelines via API (here)
Improved error handling when pipelines start -> users will get more feedback when scheduled runs fail due to for example, .yml misconfigurations
Databricks Assets - you can fetch databricks Assets using an Asset Run too
Synapse Spark Notebooks are now supported (here)
Organisation Auto-Join: for teams leveraging Okta with multiple organisations, Orchestra now supports Auto-join; you can add this as a setting here which allows users with the same email domain to join your organisational automatically
Logs are downloadable in the portal (see here)
Sigma and Tableau Lineage support (see below)
Medium 🧠
🧠 Happy New Year 2026: The Year Data + AI Stop Being Optional (link)
🧠 Catalog-Linked Databases to bring Iceberg tables in AWS Glue to Snowflake (link)
🧠 Snowflake Storage Lifecycle: How to Cut Storage Costs (link)
🧠 Towards Generalizable and Efficient Large-Scale Generative Recommenders (link)
🧠 What Advent of Code Has Taught Me About Data Science (link)
🧠 The Real Challenge in Data Storytelling: Getting Buy-In for Simplicity (link)
🧠 The “Big Data” Trap: Why Internet-Scale Solutions Are Overkill for the Enterprise (link)
🧠 Data Science Gets the Credit. Data Engineering Does the Heavy Lifting (link)
LinkedIn🕴
🕴 THE DATA WAREHOUSE BLUES (link)
🕴 Why the Data World Didn’t Have to Choose: Kimball + Inmon Both Won (link)
🕴 Why Outsourcing Everything Is the Most Dangerous Startup Strategy of 2026 (link)
🕴 2026 Is the Year AI Gets Boring (and That’s a Good Thing) (link)
🕴 The Difference Between Data Teams That Survive and Data Teams That Scale (link)
🕴 The Hardest Money You’ll Ever Earn Is Your First Dollar (link)
News 📰
📰 M&A news: Snowflake in talks to buy Observe for $1B (link)
📰 These Were The Largest Funding Rounds Of 2025 (link)
📰 AI Security market 2025 funding data, top startups, and the ServiceNow factor (link)
YouTube and Podcast 🎥
Editor’s Pick
🎥 Interview with Didier Mamma, Decathlon (link)
Special 💫
💫 Data Engineering Weekly #250 (link)
Jobs 💼
💼 Etsy Senior Data Scientist, Lifecycle Marketing at Etsy Careers (link)
💼 Senior GTM Engineer at Vibe (link)
💼 Marketing Analtics Engineer (Contract) at MetricMaven (link)
Run dbt models cheaply and easily?
If you’re looking for an easy way to run your dbt core models, look no further than Orchestra.
dbt, dbt core and dbt labs are all trademarks of dbt labs inc





Nice roundup as always. The point about AtScale getting less buzz than other modern data stack moves is spot-on. I've noticed the same thing with semantic layer plays lately. Maybe it's because the category feels less "shiny" now that everyplatform claims to have metrics layers built-in? Back when I was evaluating semantic solutions last year, it was tough to explain the ROI to execs compared to flashier acqusitions. The Snowflake investment does signal that governed context isn't optional anymore though.