What’s in today’s newsletter:
EDB acquires Altinity to merge Postgres and ClickHouse strengths🤝
Snowflake COCO boosts cloud data automation efficiency❄️
Confluent launches StreamHouse for real-time AI data⚡
Also, check out the weekly Deep Dive the EDB + Altinity deal
MULTI MODAL DATABASE
TL;DR: EDB acquired Altinity to merge Postgres and ClickHouse strengths, enhancing scalable, cost-effective open-source AI data systems that simplify infrastructure, reduce costs, and support vendor lock-in avoidance for enterprises.
EnterpriseDB acquired Altinity to combine Postgres transactional strengths with ClickHouse analytics capabilities.
The merger aims to enhance sovereign AI data systems with scalable, cost-effective, and powerful solutions.
Altinity's ClickHouse expertise helps EDB deepen open-source analytics offerings and address complex data needs.
This acquisition streamlines data infrastructure, reduces costs, and promotes vendor lock-in avoidance for enterprises.
Why this matters: EDB’s acquisition of Altinity integrates transactional and analytical database strengths, empowering enterprises with scalable, cost-efficient AI data systems. This fusion simplifies infrastructure, enhances data sovereignty, and counters vendor lock-in, aligning with modern demands for flexible, open-source solutions critical to competitive, innovative AI-driven operations.
SNOWFLAKE
TL;DR: Snowflake COCO is a new automation framework using event-driven agents to streamline and scale cloud data workflows, improve error recovery, and enable customizable, multi-cloud orchestration for enhanced operational efficiency.
Snowflake COCO is a new framework enhancing automation and orchestration in the Snowflake Data Cloud platform.
COCO uses cloud-native, event-driven agents to execute tasks and support both declarative and imperative workflows.
The framework improves operational efficiency, scalability, and error recovery in data operations through intelligent automation.
Its modular design allows integration with third-party tools, enabling customizable solutions for complex multi-cloud environments.
Why this matters: Snowflake COCO revolutionizes data operations by automating complex workflows, reducing manual effort, and enhancing scalability. Its flexible, modular design supports diverse enterprise needs and multi-cloud environments, enabling faster, more reliable data-driven decisions and aligning with the industry’s shift toward intelligent, autonomous cloud data management.
STREAMING DATA
TL;DR: Confluent's StreamHouse Lightning Tables combines streaming scalability with data lake performance to enable real-time AI data processing, multi-cloud support, and continuous model updates, boosting enterprise AI agility and decision accuracy.
Confluent introduced StreamHouse Lightning Tables to enhance real-time streaming data processing for AI applications.
The feature combines streaming scalability with data lake performance for low-latency, high-throughput data handling.
StreamHouse supports multi-cloud and seamless integration with AI workflows for continuous model updates and feature engineering.
This innovation boosts enterprise agility, accelerates AI innovation, and improves decision-making in dynamic industries.
Why this matters: StreamHouse Lightning Tables enable enterprises to process and update AI models in real-time, greatly enhancing agility and decision accuracy in competitive markets. This advance bridges streaming data and AI, accelerating innovation cycles and setting a new standard for continuous, adaptive AI model management across industries.
EVERYTHING ELSE IN CLOUD DATABASES
Top SQL Tools to Boost Query Writing & Testing
AgeDB Eyes AI & Graph Tech for Data Innovation
Denodo Leads Gartner Data Integration Quadrant 7 Years
Data Observability Boosts Banking Pipeline Accuracy
Master Data Analysis: Best Practices for AI Era
Decentralized GPUs power open data graph AI systems
Google unifies BigQuery analytics in GA
Kenya to unlock banking data, boost fintech growth
Databricks Leases Major Space at Chicago’s Franklin
Neo4j & IBM partner to boost AI with graph layers
Microsoft glitches hit Canada after clock changes end
YugabyteDB Powers Scalable Distributed SQL Solutions
DEEP DIVE
2026 is becoming the year of the agentic database and/or the multi-modal database. I am seeing a distinct trend.
There are so many examples that we are seeing such as:
EDB + Altinity: Postgres plus ClickHouse, with a shared Iceberg catalog and common access controls planned.
ClickHouse: its analytical engine plus a native Postgres service, the same pairing from the other direction.
Aurora PostgreSQL + DuckDB: an operational database querying Iceberg and Parquet in S3 directly, with DuckDB for in-process analytics.
Google AlloyDB + BigQuery: AlloyDB joins live rows with Iceberg lake data in one Postgres query.
Databricks Lakebase: a Postgres engine (from the Neon acquisition) sitting inside the lakehouse.
Snowflake Postgres: a Postgres engine (from the Crunchy Data acquisition) alongside the warehouse, with Spark workloads now being pulled in too.
Microsoft Fabric: SQL database plus warehouse, lakehouse and real-time analytics over OneLake.
Confluent Streamhouse: a stream that can be queried like a table.
In many sources that I have read over the past few days, the EDB + Altinity news appears to be the most consequential database sales and announcements of the year.
Here is a brief synopsis of the deal and what EDB is trying to accomplish from a technical standpoint:
EnterpriseDB (EDB) has acquired Altinity, bringing together the commercial stewards of the two most dominant open-source database ecosystems: PostgreSQL for mission-critical transactional workloads and ClickHouse for hyper-scale, sub-second columnar analytics. While financial terms were undisclosed, the strategic intent is unmistakable.
EDB is integrating Altinity’s deep engineering bench, which accounts for 888 merged ClickHouse pull requests and the gold-standard Altinity Kubernetes Operator, into its EDB Postgres AI platform. Crucially for enterprises wary of licensing turbulence, EDB committed that Altinity’s core deliverables, including Altinity Stable Builds, Project Antalya (its object-storage integration branch), and the operator, will remain fully open source under Apache 2.0 alongside ongoing upstream contributions.
From an architectural standpoint, the deal aims to dismantle the fragile, high-maintenance "glue code" that enterprises traditionally maintain between operational rows and analytical columns.
Instead of managing external Change Data Capture (CDC) pipelines, Kafka topics, and staging layers to mirror OLTP state into an external warehouse, EDB is positioning PostgreSQL as the transactional "ledger" and ClickHouse as the analytical "lens". The near-term technical roadmap centers on establishing a shared Apache Iceberg catalog, unified access controls, and zero-copy queries across both engines.
This design allows ClickHouse to query the live data Postgres writes in under a second while running under a consistent security and governance perimeter.
For data architects and engineering leaders, this marks a decisive turn away from fragmented, single-model database sprawl. Where proprietary cloud vendors are nudging customers toward proprietary HTAP runtimes or closed compute meters, EDB and Altinity are building a sovereign, multi-modal alternative.
It delivers ACID relational transactions, native vector retrieval via pgvector, and petabyte-scale columnar aggregations in a stack that can run anywhere, on-premises, across bare-metal Kubernetes, or in hybrid multi-cloud environments.
In an agentic era where autonomous workflows demand immediate access to fresh operational context without waiting for batch ETL, uniting the transactional record with real-time analytical compute under one open-source roof is the clearest sign yet that the database platform is collapsing the pipeline into the engine.
By bringing open-source OLTP and real-time OLAP under a single commercial and operational roof, EDB and Altinity are dismantling the fragile, expensive integration pipelines that long separated transactional and analytical estates.
For enterprise architects, this deal marks a defining pivot toward unified, sovereign multi-modal data platforms capable of serving real-time agentic workloads without the penalty of proprietary cloud lock-in.
Gladstone Benjamin
