Amazon Accelerates Supply Chain Decision Making by Implementing an Innovative Analytics Architecture using Dremio
The Agentic Lakehouse
The only lakehouse built for agents, managed by agents
TRUSTED BY:
AI CAPABILITIES
AI Capabilities That Drive Business
Insights and Innovation
Dremio is designed to make every business user an expert analyst. There is no compromise between simplicity, low costs, and lock-in. Dremio delivers instant insights for every team.
1
2
3
-
Unified Access for Agents
Give Agents Governed Access to the Data they Need
- Zero-ETL Query Federation
- Process all data types: structured, semi-structured, and unstructured
- Fine-grained access controls ensure agents only access the right data
-
AI Semantic Layer
Provide Agents the Business Context they Need to Deliver High Quality Answers
- Semantic Search helps agents find the right data
- AI-Gen Wikis and Labels create the business metadata agents need
- Semantic metadata is stored in an open catalog
-
Fastest Lakehouse
Get Fast Answers from your Agent
- 20x Faster than Cloud Data Warehouses
- Apache Arrow Columnar Processing
- Performance is Managed Autonomously
AUTONOMOUS
Autonomous Management
Dremio’s platform learns from every query, automatically creates optimizations, and rewrites SQL in real-time to deliver consistent sub-second performance. No manual tuning, no performance engineers, no waiting—just really fast performance at the lowest cost.
Enhance Semantics
When prompted, Dremio enhances the semantic layer by automatically generating Wikis and Labels for Tables and Views.
Materialize
Dremio analyzes query patterns to automatically produce materializations (aggregations, joins, etc.) that optimize both the performance and cost of your queries. These materializations are kept live as the source tables update.
Query Rewrite
When agents query views, Dremio autonomously reviews the query, looks for optimized materializations (Reflections), and rewrites the queries on the fly to deliver 20x faster queries at a fraction of the cost of cloud data warehouses.
Cluster
Dremio maintains optimal data layouts, intelligently reorganizing data based on clustering keys and eliminating the need for traditional partitioning.
LEADERS IN OPEN SOURCE
The Lakehouse OSS
Standards
From the original co-creators of Apache Polaris and Apache Arrow, Dremio is the only lakehouse built natively on Apache Iceberg, Polaris, and Arrow - providing flexibility, preventing lock-in, and enabling community-driven innovation.
Apache Arrow Dremio Co-created
Lakehouse in-memory and transport format (100M+ downloads/month)
Apache Iceberg Key contributor and leading educator
Lakehouse table format
Apache Polaris Dremio Co-created
Lakehouse metadata catalog
CASE STUDIES
Fueling AI innovation across industries
1000s of companies across all industries trust Dremio.
Empowering our customers with robust data governance on STACKIT is crucial. Dremio's Open Catalog provides a secure central platform to manage data and ensures full data sovereignty within the trusted STACKIT Cloud
Benjamin Schweizer Senior Manager Domain Product Owner STACKIT
ABC Supply delivers self-service analytics using Dremio to organize domain data and accelerate development
Stephen Brehm Manager of Business Intelligence ABC Supply
Maersk built their next generation data platform for unified analytics on Dremio making it the go-to place for end-user SQL queries in only six months
Québec Blue Cross streamlined data product delivery process with full CI/CD by leveraging Dremio with DBT
Nutanix fast tracked data provision to key business verticals by building Data-as-a-Service platform with Dremio
USE CASES
Modernize Your Data Infrastructure
See real-world use cases of companies preparing their data for AI and innovation.
AI-Driven Analytics
Enable AI agents with direct access to enterprise data through natural language. MCP Server provides zero-integration connectivity to LLMs and AI frameworks.
Unify Data & Analytics
Eliminate data silos with a unified semantic layer and query federation that provides consistent definitions, metrics, and business logic across all your data sources.
Warehouse to Lakehouse
Get faster performance, more flexibility, and lower management overhead than traditional warehouses. The best warehouse is an Iceberg Lakehouse.
Lake to Iceberg Lakehouse
Connect on-premises and cloud data lakes into a unified lakehouse architecture through advanced query federation and virtual data integration - no ETL required.
Hybrid Lakehouse
Connect on-premises and cloud data lakes into a unified lakehouse architecture through advanced query federation and virtual data integration - no ETL required.
Data Fabric
Connect disparate data sources across hybrid and multi-cloud environments with a unified architecture that enables consistent governance, discovery, and access controls.
COMMUNITY
Our Community
Our community drives AI innovation and inspires us every day. From success stories to real-world impact, here’s how people are experiencing Dremio
Follow us
Yesterday, I went to Data Engineering Meetup at the Dremio office, and I had a great time, even though I arrived late. My apologies! I really enjoyed the speakers and all their talks, including: - Ashley Farrugia: Developing AI-Ready Data Products with Dremio and dbt...
From empowering enterprises to democratize data access to supporting AI-driven innovation, Dremio truly stands out as a game-changer in the industry. I especially loved hearing about real-world examples, like how one company reduced time-to-insights by 90%!
Today, I explored how Apache Iceberg handles data changes like updates and merges, manages snapshots, and optimizes data access with partitioning, all without rewriting entire datasets a true game-changer for scalable data management.
Thank you, Dremio, for an evening full of great conversations, delicious food, and amazing company! 🙌🍷
Switching from Redshift to Dremio can make your analytics faster, simpler, and more cost efficient. If you own the platform or make the call, book a November meeting with the Dremio team to explore a modern lakehouse built for AI. There is a perk.
Everyone’s buzzing about the AI boom but I see a lot of teams stuck with legacy data warehouse architectures that simply can’t keep up. If that sounds like you, what can you do about it? 👇 Organizations need open and flexible platforms to support today’s modern data/AI workloads. Have you stopped for a second to consider if the tech you’re using is enough? Case in point, who here is still running on Amazon Redshift? I call it the Redshift rut 😂 Redshift was great years ago. But data volumes have grown massively and workloads are nothing like they were, so legacy foundations are showing their limits.
RESOURCES
Blogs
View all
Handling Complex Data Types in Dremio
Overview Dremio provides out-of-the-box methods of handling complex data types in, for example JSON and parquet datasets. Common characteristics are embedded “columns within columns” and “rows within columns”. ...
No articles found.
FAQs
Frequently Asked Questions
Find quick answers to common questions about Dremio's platform and its features.
Dremio continuously analyzes query patterns and automatically creates Autonomous Reflections (smart materializations), applies Iceberg Clustering (intelligent data layout), and optimizes tables—all without human intervention. The system learns and adapts as your workload changes.
The MCP Server enables AI agents to discover and use data tools like RunSqlQuery and GetSchemaOfTable automatically. AI-enabled semantic search lets you find data using plain language. No custom integrations required—just natural language access to your enterprise data.
No data movement required. Dremio queries data where it lives using advanced federation. Your existing SQL queries work unchanged—they’re automatically optimized at runtime to take advantage of Reflections and intelligent caching. For Iceberg datasets, the management of Reflection based acceleration will be autonomous.
Yes, Dremio can be used with Python. Thanks to Dremio’s support for REST, ODBC, JDBC, and Apache Arrow Flight interfaces, you can connect to Dremio from virtually any programming language, including Python. Dremio also offers Python libraries such as dremio-simple-query and pyDremio, which make it easy to query and analyze data directly from your Python environment.
Yes. Dremio provides an ARP (Advanced Relational Pushdown) framework that allows you to build custom connectors. The community has already created connectors for many data sources using this framework. However, when possible, we recommend ingesting data into Apache Iceberg tables instead. This allows you to take full advantage of Dremio’s Autonomous Performance features, including acceleration, optimization, and caching.
Traditional systems require manual partitioning schemes that become maintenance nightmares. Dremio’s Iceberg Clustering automatically organizes data for optimal performance without any manual partition management. You get better performance with zero partitioning overhead.
All your current tools work exactly the same. Dremio provides one-click integrations with Power BI, Tableau, and other BI platforms. Users see faster dashboards and queries without changing anything in their workflow.
Yes. The Dremio Catalog is built on Apache Polaris, which implements the Apache Iceberg REST Catalog specification. This means any tool that supports the Iceberg REST catalog—such as Spark, Flink, and others—can read from and write to tables managed by the Dremio Catalog. Likewise, Dremio can connect to external Iceberg catalogs using its REST Catalog connector.
Yes. Dremio’s OPTIMIZE and VACUUM commands can be used on any Iceberg table that Dremio has write access to, regardless of which catalog it resides in. However, automatic optimization features are currently only available for Iceberg tables managed within the Dremio Catalog.
