Member-only story

Materialize Once, Query Anywhere: Amazon Redshift's New Iceberg Materialized Views

How a new AWS capability lets data engineers compute expensive analytics once and share the results across Redshift, Athena, Spark, and other Iceberg-compatible engines—without maintaining duplicate datasets.

5 min readOct 23, 2026

Imagine your analytics team calculates daily revenue using Amazon Redshift. Your data science team independently calculates similar metrics in Spark, while analysts run their own aggregations in Athena.

Friend link for non-members- Here

Three engines. Three implementations. Three opportunities for inconsistent results.

Now imagine computing those metrics once and publishing them as an open table that every team can query.

On October 5, 2026, AWS announced support for creating and refreshing Apache Iceberg materialized views in Amazon Redshift. These views store precomputed results as standard Iceberg tables in Amazon S3 or Amazon S3 Tables, registered in the AWS Glue Data Catalog. Compatible engines can consume the results without copying the dataset into their own storage formats.

Towards Data Engineering

Materialize Once, Query Anywhere: Amazon Redshift's New Iceberg Materialized Views

How a new AWS capability lets data engineers compute expensive analytics once and share the results across Redshift, Athena, Spark, and other Iceberg-compatible engines—without maintaining duplicate datasets.

Written by Mohit Daxini

Published in Towards Data Engineering

No responses yet

More from Mohit Daxini and Towards Data Engineering

From Natural Language to ETL Pipelines: How AI Is Changing AWS Data Engineering

Generative AI can speed up data pipeline development, but experienced data engineers still need to own the architecture, correctness, and…

6 SQL Concepts Every Data Engineer Should Know

Many years ago, I was having a conversation with a friend who was looking to break into data engineering.

I Ask Every Candidate the Same SCD2 Follow-Up. Most of Them Go Quiet.

Every candidate has the definition ready. It arrives fast and clean, like a card being pulled.

Amazon Bedrock Just Made RAG Knowledge Bases Easier to Maintain: What Data Engineers Should Know

AWS adds scheduled synchronization to Managed Knowledge Bases, reducing custom orchestration while introducing important decisions around…

Recommended from Medium

AWS Just Put DuckDB Inside PostgreSQL. Data Engineers Have Some Rethinking to Do.

If you’ve built more than a couple of data pipelines, you’ve built this one before. Your orders live in Postgres. Three years of history…

Building an Agentic Knowledge Graph for Billions of Events

Rebuilding Netflix-inspired observability with AI

The Most Dangerous Data Engineering Bug Isn’t a Failed Pipeline. It’s a Pipeline That Runs Twice.

How to build retry-safe, incremental data pipelines using Python and PostgreSQL — without duplicate records, corrupted metrics, or…

How do we implement SCD-Type 2?

SCD Type 1 — Overwrite

I Turned a 30‑Step Spark Pipeline Into a Single YAML File

A few months ago, we had a Spark pipeline that had become difficult to change.

Apache Spark 4.0 to 4.2: What’s New, What’s Changed, and What Data Engineers Should Know

In modern data engineering, Apache Spark sits at the center of almost every Big data pipeline. For a long time most of us were comfortable…