Dremio

What Dremio does

Dremio is the Agentic Lakehouse, an Iceberg-native data platform designed for agents and managed by agents. It provides instant, governed access to enterprise data for knowledge workers and AI agents through any LLM or tool. The platform enables federated queries across any data source without ETL pipelines and includes an AI Semantic Layer that adds business context for a single source of truth. Dremio autonomously manages clustering, optimization, and compaction, delivering trusted insights without infrastructure complexity. It is an open, high-performance data lakehouse platform built to accelerate AI and analytical workloads across all enterprise data, available as a fully managed cloud service or self-managed software. Dremio is a lead contributor to Apache Iceberg and co-creator of Apache Arrow and Apache Polaris.

Products

• Dremio CloudA fully managed platform with zero infrastructure management, automatic updates, scaling, and optimization. Offers instant setup and automatic feature releases. Supported on AWS (Azure coming soon).
• Dremio EnterpriseA self-hosted deployment option offering complete infrastructure control for security, compliance, and data policies. Flexible deployment on Cloud (AWS, Azure, GCP), Kubernetes, or On-premises, or as Dremio-as-a-Service (DaaS). Allows custom integration with existing enterprise infrastructure, authentication, and monitoring tools.
• Dremio Community EditionA free query engine that can be deployed self-managed on-premises or in a preferred cloud environment.
• AI AgentAn integrated AI agent that allows users to ask natural-language questions, dynamically generates SQL, validates results, and delivers insights alongside visualizations.
• AI Semantic LayerA unified business context layer that provides the business and technical context agents need to interpret data correctly, surfaces metadata, auto-generates documentation and labels, and enables semantic search for trusted datasets.
• Intelligent Query EngineA high-performance SQL analytics engine built on Apache Arrow, optimized for workloads, supporting federated queries across object storage, relational databases, and NoSQL systems without ETL.
• Open Catalog (Apache Polaris)Manages metadata for Iceberg tables in the lakehouse, including schemas, table definitions, and query metadata, enabling unified governance and fine-grained access control across data assets.
• Autonomous ReflectionsOptimizes performance without manual tuning by analyzing workloads and creating Reflections when beneficial, rewriting queries to use Reflections, and maintaining table health through background processes like compaction and metadata cleanup.

Key capabilities

• Unified Data & Semantics
• Federated queries across any data source without moving data
• Query structured, semi-structured, and unstructured data
• Iceberg-Native Platform
• Arrow-based query engine processes Iceberg data natively
• Open Catalog (Apache Polaris) enables read/write across any Iceberg REST engine
• Automatic Iceberg table optimization (clustering, compaction, vacuum, variant shredding)
• MCP integration connects Claude, ChatGPT, Gemini, and more
• Dremio CLI for coding agents like Claude Code and Codex
• Built-in AI agent
• Autonomous Reflections observe query patterns and accelerate queries
• Serverless architecture scales on demand
• End-to-end caching reduces redundant compute and shortens response times
• AI agents inherit user identity and access controls automatically
• Fine-grained and role-based access controls enforced from client to source
• OAuth tokens flow through credential vending to every data source
• End-to-End Access Control
• High-performance SQL analytics built on Apache Arrow
• Supports federated queries across object storage, relational databases, and NoSQL systems
• Unified analytics across all data sources without ETL
• Manages metadata for Iceberg tables
• Optimizes performance without manual tuning
• Handles compaction, metadata cleanup, and other automated maintenance tasks
• Zero Infrastructure Management (Dremio Cloud)
• Instant Setup (Dremio Cloud)
• Automatic Feature Releases (Dremio Cloud)
• Complete Infrastructure Control (Dremio Enterprise)
• Flexible Deployment (Dremio Enterprise)
• Custom Integration (Dremio Enterprise)
• Git-like version control for data management
• Automatic data optimization for high performance analytics
• Natural-language analytics using Dremio’s AI agent and external agents
• Governed metadata to understand data
• Auto-generates documentation and labels
• Semantic search for easy data discovery
• Apache Arrow-native engine queries Iceberg and Parquet directly
• Apache Arrow Flight and ADBC for high-throughput data transfer
• Autonomous Reflections auto-creates and manages materializations
• Optimizer rewrites incoming queries to use pre-computed results transparently
• Reflections refresh automatically as base tables update
• Query structured, semi-structured, and unstructured data natively across 35+ source types
• Optimized pushdowns to delegate filters, projections, aggregations, and joins
• Full INSERT, UPDATE, DELETE, MERGE, and TRUNCATE operations on Apache Iceberg tables
• Time-travel queries via AT SNAPSHOT and AT TIMESTAMP syntax
• Native support for Apache Polaris and other open catalogs
• Serverless Elastic Engines scale to zero when idle and scale out on demand
• Workload isolation prevents resource contention
• Workload management and queue controls for predictable SLAs
• Automated Data Integration & Virtualization
• Unified fabric connects all data sources in one virtualized layer
• Real-time access eliminates complex ETL
• Central governance for access, lineage, compliance
• Role-based controls and audit trails for governed analytics
• AI-Ready Data Discovery & Self-Service Analytics
• Intelligent search for trusted datasets
• Semantic layer accelerates innovation and AI adoption
• Zero-ETL Federation Across All Enterprise Data
• Universal data access provides real-time virtualization across cloud, on-premises, and SaaS systems
• Self-Service Analytics for the AI Era
• Enterprise Security & Governance for Agentic AI
• Fine-grained controls, comprehensive audit trails, and centralized metadata management powered by Apache Polaris
• SOC II Type 2 certification
• ISO 27001:2013 certification
• GDPR compliance
• CCPA compliance
• End-to-End Encryption (TLS 1.2+ in Transit and AES-256 at Rest)
• Customer-Managed Encryption Keys
• Connection via PrivateLink
• Role-based Access Control (RBAC)
• Fine-Grained Access Control
• Hierarchical Permission and Inheritance
• Single Sign-On (SSO), Multi-Factor Authentication (MFA)
• Identity Provider Integration (Okta, Microsoft Entra ID, Google Identity)

Use cases

• Accelerate AI and analytical workloads
• Provide instant, governed access to enterprise data for knowledge workers and AI agents
• Federate queries across any data source without ETL pipelines
• Add business context with an AI Semantic Layer
• Autonomous management of data lakehouse operations (clustering, optimization, compaction)
• Speed up decisions on billions of records (e.g., Amazon's SCOT Finance Analytics)
• Offload repetitive, expensive dashboard and reporting queries from Redshift, Snowflake, or Databricks
• Modernize legacy Hadoop workloads
• Build a distributed data architecture (Data Mesh)
• Standardize on an open data architecture (Data Lakehouse)
• Accelerate AI experimentation and reproducibility
• Natural-language analytics
• High-performance SQL analytics
• Unified metadata and governance
• Performance optimization without manual tuning
• Zero-maintenance, cloud-native experience (Dremio Cloud)
• Full control for on-prem requirements or strict compliance (Dremio Enterprise)
• Fueling AI innovation across industries
• Improved Supply Chain with >10x Query Performance and 90% Faster Project Delivery
• Customer 360 Analytics Projects
• Gen AI acceleration
• Connecting AI agents and analytics tools to an intelligent query engine, semantic layer, open catalog, and governed data sources
• Running 100+ concurrent forecasting models (Shell)
• Providing self-service access to genomic data for researchers (Genomics England)
• Transforming capital markets operations with an agentic lakehouse (World Bank Group)
• Fraud detection, risk, and customer insights in financial services
• Unifying transaction, KYC, market, and regulatory data in financial services
• Real-time feature lookup for fraud models
• Creating a single version of risk metrics for Basel III, FRTB, IFRS 9, Solvency II programs
• Unifying customer identity and behavioral data for next-best-action, personalized offers, and real-time servicing
• Providing portfolio managers, quants, and distribution teams with intraday portfolio, performance, and risk views
• Unifying transaction monitoring, watchlist screening, customer due diligence, and data lineage for compliance
• Unifying product telemetry, CRM, feature store, and operational data in technology companies
• Fastest Path to Product and Customer Analytics (technology)
• Querying product, customer, and event data in place with zero ETL (technology)
• Enforcing governed access, full lineage, and audit-ready compliance across multi-tenant platforms (technology)
• Product and data teams querying petabyte-scale event telemetry, session data, and feature flags alongside CRM and support signals
• Revenue operations and customer success teams unifying product usage, CRM, support tickets, and billing data for account health scoring, churn prediction, and expansion signal detection
• ML engineers and data scientists building and serving training datasets and feature stores directly on open Iceberg tables
• Data platform and infrastructure teams querying data across AWS, Azure, and GCP environments
• Migrating workloads from Snowflake or BigQuery to an open Iceberg lakehouse
• Warehouse to Lakehouse Migration
• Data Unification with Immediate Value
• Warehouse Cost Optimization
• Open Architecture and Future-Proof Flexibility
• Unified Analytics for All Workloads
• Automated table optimization (clustering, compaction, vacuum)
• AI-enabled semantic search
• Zero-ETL Federation Across All Enterprise Data
• Self-Service Analytics for the AI Era
• Streamlining ETL processes and reducing data engineering workloads (Amazon)
• Providing reliable, consistent, and accurate insights to internal users (Amazon)
• Processing enormous volumes of data (Amazon)
• Creating financial reporting for supply chain systems (Amazon)
• Unifying historian, ERP, MES, and supply chain data in manufacturing
• Fastest Path to Manufacturing AI and Analytics
• Querying operational and enterprise data in place with zero ETL (manufacturing)
• Enforcing fine-grained access, full lineage, and audit-ready governance across regulated workloads (manufacturing)
• Manufacturing and operations teams connecting historian telemetry with maintenance history and work orders to identify equipment failure patterns
• S&OP and supply chain teams unifying demand signals, inventory, supplier lead times, and logistics data across multiple ERP instances
• Quality and engineering teams correlating vision inspection results, SPC data, and material lot traceability to identify defect root causes
• Operations teams getting real-time OEE reporting across a multi-plant footprint
• ML engineers and planning teams building demand forecasting models on unified historical sales, inventory, supplier lead times, and external signals
• Unifying EDC, LIMS, omics, EHR, and commercial data in life sciences and healthcare
• Fastest Path to Life Sciences AI and Analytics
• Querying clinical and research data in place with zero ETL (life sciences)
• Enforcing governed access, full lineage, and audit-ready compliance across regulated data (life sciences)
• Clinical data management teams connecting EDC, LIMS, biomarker, and imaging data for interim analyses and NDA submissions
• Research bioinformaticians querying multi-omics datasets alongside clinical phenotype data
• RWE and health economics teams unifying claims, EHR, specialty pharmacy, and patient registry data for post-market studies, label expansions, and payer submissions
• Drug safety teams correlating adverse event reports with product exposure data for signal detection and expedited FDA reporting
• Brand analytics and market access teams unifying prescriber, payer, specialty pharmacy, and field force data for launch tracking, formulary access monitoring, and pull-through analytics
• Unified Data Fabric to access distributed data
• Automate integration across on-prem, cloud, and SaaS for a unified, real-time, governed view
• Streamlining analytics by connecting diverse data sources (TD Securities)
• Standardizing data access and governance for sensitive customer data (TransUnion)
• Accelerating critical analytics, reducing query times and management overhead (FactSet)
• Improving operational agility and decision-making by enabling self-service analytics and easier discovery of data assets (ABC Supply Co.)
• Building a next generation data platform for unified analytics (Maersk)
• Migrating SAP reporting workloads to Dremio (Maersk)

AI approach

Dremio provides an Agentic Lakehouse platform designed to accelerate AI and analytical workloads. It features an integrated AI Agent for natural-language analytics, an AI Semantic Layer to provide business context and unify data, and autonomous management capabilities that optimize performance without manual tuning. Dremio connects to various AI agents and LLMs via its open MCP server and CLI, and supports custom agent frameworks. The platform is built on open standards like Apache Iceberg and Apache Arrow, and aims to provide governed, real-time access to enterprise data for both human analysts and AI agents.

Tech named: Apache Iceberg, Apache Arrow, Apache Polaris, MCP (Model Context Protocol), Claude, ChatGPT, Gemini, Claude Code, Codex, LangChain, LlamaIndex, REST API, Apache Arrow Flight, LLVM, C3 columnar cloud cache, Elastic Engines

Industries served

• Software Development
• Financial Services
• Technology
• Manufacturing
• Life Sciences
• Healthcare
• Energy
• Logistics
• Retail
• Public Safety
• Food Services
• Investment
• Government

What it says sets it apart

• The only Iceberg-native data platform built for agents and managed by agents
• Federated queries reach any source without ETL pipelines
• AI Semantic Layer adds business context for a single source of truth
• Lakehouse manages itself, running clustering, optimization, and compaction autonomously
• Lead contributor to Apache Iceberg and co-creator of Apache Arrow and Apache Polaris
• Open, high-performance data lakehouse platform built to accelerate AI and analytical workloads
• Connects any AI agent or BI tool in minutes
• Scales and optimizes itself without manual tuning
• Enforces access controls end to end so permissions travel with your data
• Combines low cost and flexibility of a data lake with performance and governance of a data warehouse
• Runs queries directly on open formats in object storage, eliminating data copies, proprietary storage fees, and constant ingest pipelines
• Automatic query acceleration
• Agentic Interface enables natural-language analytics
• MCP Server for instant access across tools and models
• Autonomous Reflections optimize performance without manual tuning
• Open Catalog (Apache Polaris) for unified metadata and governance
• Zero Infrastructure Management with Dremio Cloud
• Complete Infrastructure Control with Dremio Enterprise
• Built for teams running AI at enterprise scale
• Connect any agent in minutes with one-click integrations and hosted MCP server
• One source of truth across every system with AI Semantic Layer
• Governed, fast answers at any scale with automated acceleration and built-in governance
• Queries across 31+ connected sources simultaneously with no data movement and no manual joins required
• Every result comes back with the same row-level permissions, column masking, and business logic
• Uses open Model Context Protocol (MCP) rather than proprietary integrations, avoiding vendor lock-in
• AI Semantic Layer spans the entire federated data estate, unlike Databricks metric views or Snowflake semantic views
• Iceberg-Native from the Ground Up, not adapted or bolted on
• Sub-Second Performance at a Fraction of the Cost with autonomous optimization and Arrow-native query engine
• Governed Access for Every Tool and Agent with consistent access layer
• Automated Table Optimization (clustering, compaction, vacuum) on autopilot
• Decoupled storage and compute delivers up to 20x performance improvement at data lake cost
• AI Semantic Layer generates wikis and labels for every dataset automatically
• Apache Arrow-native engine queries Iceberg and Parquet directly, no translation layer or format conversion
• Apache Arrow Flight and ADBC eliminate serialization overhead for high-throughput data transfer
• Autonomous Reflections auto-creates and manages materializations based on observed query patterns
• Optimized pushdowns delegate filters, projections, aggregations, and joins to each source engine
• Full DML support (INSERT, UPDATE, DELETE, MERGE, TRUNCATE) on open Apache Iceberg tables
• Time-travel queries via AT SNAPSHOT and AT TIMESTAMP syntax
• Serverless Elastic Engines scale to zero when idle and scale out on demand
• Workload isolation prevents resource contention
• Unified Data Fabric automates integration across on-prem, cloud, and SaaS for a unified, real-time, governed view without complex ETL or data movement
• Enterprise Security & Governance for Agentic AI with fine-grained controls, comprehensive audit trails, and centralized metadata management powered by Apache Polaris
• Certified Security (SOC II Type 2, ISO 27001:2013, GDPR, CCPA)
• End-to-End Encryption (TLS 1.2+, AES-256)
• Customer-Managed Encryption Keys
• Connection via PrivateLink
• Role-based Access Control (RBAC), Fine-Grained Access Control, Hierarchical Permission and Inheritance
• Single Sign-On (SSO), Multi-Factor Authentication (MFA), Identity Provider Integration