Datastrato Iceberg Lakehouse Platform

datastrato.ai

Free trial Open source LinuxAPISelf-hosted

Spec score 8.4 #4 of 23 in Data Lakehouse Platforms

What it is

Datastrato Iceberg Lakehouse Platform is a federated metadata catalog for connecting data sources and governing access across clients. Built on Apache Gravitino and Iceberg REST Catalog, it provides a standards-compliant REST catalog with metadata caching and table maintenance for production lakehouses. It can register Iceberg, Hive, Glue, relational databases, Kafka, filesets, and models in place without copying metadata. Spark, Flink, Trino, and Dremio are named Iceberg catalog clients, and the same namespace resolves across BigQuery. Governance includes catalog-level role-based access, row filtering, column masking, tag-based policies, and auditing across engines and agents. S3, GCS, and ADLS are supported with short-lived scoped credentials per principal. Agents can use an MCP metadata interface and GVFS under the same access controls. Per-table compaction, snapshot expiry, and orphan cleanup jobs can be sent to Airflow, Kubernetes, EMR, or Databricks. Enterprise is installed with a Helm chart in Kubernetes and activated by license key. Pricing is on request; the commercial distribution includes support with a defined response commitment and maintained release line.

Spec sheet

What it does: Datastrato provides a federated metadata catalog and Iceberg REST platform for governing data and AI assets across clouds, engines and formats.Read 14 Oct 2026 · datastrato.ai
Purpose: Datastrato is a federated metadata catalog built on Apache Gravitino and Iceberg REST Catalog for connecting data sources and governing access across clients.Read 5 Oct 2026 · datastrato.ai
Audience: Datastrato says its platform is for enterprise teams and developers working with metadata management, governance, and AI-ready infrastructure.Read 5 Oct 2026 · datastrato.ai
Deployment: Datastrato Enterprise is installed from a Helm chart into a Kubernetes cluster and activated with a license key; local evaluation uses local Kubernetes.Read 5 Oct 2026 · datastrato.ai
Integrations: The product page names Spark, Flink, Trino, and Dremio as Iceberg catalog clients, and says the same namespace also resolves across BigQuery.Read 5 Oct 2026 · datastrato.ai
Support: Standard and Premium support include a dedicated Slack Connect channel, named technical account manager and one-hour P0 response, with 24/7 response on Premium.Read 14 Oct 2026 · datastrato.ai
License: Apache Gravitino, the underlying engine, is licensed under Apache License 2.0; Enterprise is a commercial distribution.Read 14 Oct 2026 · github.com
Founded: LinkedIn lists Datastrato's founding year as 2023.Read 14 Oct 2026 · linkedin.com
AI access: Agents can use an MCP metadata interface and GVFS for files under the same RBAC, FGAC, and audit controls as other clients.Read 5 Oct 2026 · datastrato.ai
AI agents: The product exposes governed metadata through MCP and unstructured files through GVFS, with the same policy and audit path used by human clients.Read 14 Oct 2026 · datastrato.ai
Cloud storage: Native storage support covers Amazon S3, Google Cloud Storage and Azure Data Lake Storage through URI-based backend dispatch.Read 14 Oct 2026 · datastrato.ai
Community history: A DEV Community introduction says Gravitino was initiated by Datastrato, open-sourced in 2023, donated to Apache Incubator in 2024 and graduated to an Apache Top-Level Project in May 2025.Read 14 Oct 2026 · dev.to
Company headquarters: LinkedIn lists Datastrato headquarters in San Mateo, California.Read 14 Oct 2026 · linkedin.com
Company size: LinkedIn lists Datastrato as a privately held software-development company with 11–50 employees.Read 14 Oct 2026 · linkedin.com
Compliance claims: Datastrato states SOC 2 Type II, ISO 27001 readiness, HIPAA BAA and GDPR alignment, with audit reports available under NDA.Read 14 Oct 2026 · datastrato.ai
Credential vending: Short-lived scoped credentials are minted per principal for AWS, GCP and Azure storage access.Read 14 Oct 2026 · datastrato.ai
Deployment requirement: Datastrato Enterprise is deployed as a Helm chart into Kubernetes and activated with a license key.Read 14 Oct 2026 · datastrato.ai
Documentation depth: The Enterprise documentation is organized into installation, metadata exploration, governance, automation, source connections, authorization, monitoring, configuration, engines, agents, reference and release notes.Read 14 Oct 2026 · datastrato.ai
Enterprise identity: Datastrato Enterprise adds SCIM provisioning, local fallback accounts and RBAC across federated catalogs.Read 14 Oct 2026 · datastrato.ai
Enterprise lineage: Enterprise documentation states that it records metadata-operation audits and collects lineage from Spark through OpenLineage.Read 14 Oct 2026 · datastrato.ai
Enterprise operations: The commercial distribution adds a supported Helm chart, health and readiness endpoints, service metrics and an administration UI.Read 14 Oct 2026 · datastrato.ai
Enterprise support: Datastrato Enterprise includes support with a defined response commitment and a maintained release line.Read 5 Oct 2026 · datastrato.ai
Federated catalog: It registers Iceberg, Hive, Glue, relational databases, Kafka, filesets and models in place under one namespace without copying metadata.Read 14 Oct 2026 · datastrato.ai
Federation: It can register Iceberg, Hive, Glue, relational databases, Kafka, filesets, and models in place without copying metadata.Read 5 Oct 2026 · datastrato.ai
Founder: Datastrato's origin-story page identifies Junping Du as founder and CEO.Read 14 Oct 2026 · datastrato.ai
GitHub activity: The Apache Gravitino repository shows about 3.2k stars and 954 forks.Read 14 Oct 2026 · github.com
Governance: RBAC, row filtering, column masking and tag-based policies are evaluated at the catalog for every request.Read 14 Oct 2026 · datastrato.ai
High availability: Metadata persistence can use PostgreSQL or MySQL through a pluggable JDBC backend with disaster-recovery support.Read 14 Oct 2026 · datastrato.ai
Iceberg platform: Its Iceberg Lakehouse Platform provides a standards-compliant Iceberg REST Catalog with metadata caching and table maintenance for production lakehouses.Read 5 Oct 2026 · datastrato.ai
Iceberg REST: The platform implements a standards-compliant Iceberg REST Catalog with hierarchical namespaces and three-level catalog.namespace.table naming.Read 14 Oct 2026 · datastrato.ai
Independent API detail: The Iceberg Knowledge Base shows PyIceberg connecting to Gravitino's REST endpoint and using standard list-tables and load-table operations.Read 14 Oct 2026 · iceberglakehouse.com
Independent criticism: AI Tech Suite says Datastrato focuses on metadata and therefore requires external tools for physical data movement.Read 14 Oct 2026 · aitechsuite.com
Independent description: The Apache Iceberg Knowledge Base describes Gravitino as an open-source meta-catalog with one REST API spanning Hive, Iceberg REST, JDBC and fileset backends.Read 14 Oct 2026 · iceberglakehouse.com
Independent setup limitation: AI Tech Suite says initial setup and integration require substantial technical expertise.Read 14 Oct 2026 · aitechsuite.com
MCP SDK integration: Datastrato publishes an Apache-2.0 Python MCP server using FastMCP and uv, with catalog, schema, table, tag, role and model tools.Read 14 Oct 2026 · github.com
Metadata cache: Datastrato claims its Iceberg REST metadata cache delivers four-times-faster reads under concurrent AWS load.Read 14 Oct 2026 · datastrato.ai
Named engines: Datastrato names Apache Spark, Trino, Flink, Dremio, BigQuery, Ray and Daft as clients or engines.Read 14 Oct 2026 · datastrato.ai
Named integrations: Documented integrations include Airflow, Kubernetes, EMR, Databricks, Kafka, Splunk, Datadog, Microsoft Sentinel, LDAP and Active Directory.Read 14 Oct 2026 · datastrato.ai
Open-source installation: Apache Gravitino can be downloaded as a binary, started with gravitino.sh, run through Docker or built from source with Gradle.Read 14 Oct 2026 · github.com
Operations: The product page describes per-table compaction, snapshot expiry, and orphan cleanup jobs dispatchable to Airflow, Kubernetes, EMR, or Databricks.Read 5 Oct 2026 · datastrato.ai
Production adopters claim: The DEV Community introduction names Xiaomi, Tencent, Zhihu, Uber and Pinterest as production users of Gravitino.Read 14 Oct 2026 · dev.to
Public API: The platform retains Apache Gravitino REST endpoints, client libraries and connector names, and exposes standard Iceberg REST APIs.Read 14 Oct 2026 · datastrato.ai
Release status: Apache's release catalog lists Gravitino 1.3.1 released on 30 September 2026.Read 14 Oct 2026 · release-catalog.apache.org
Security and compliance: Datastrato's homepage states SOC 2 Type II, ISO 27001 readiness, HIPAA BAA, and GDPR alignment; audit reports and control documentation require NDA-gated access.Read 5 Oct 2026 · datastrato.ai
Security posture: The maker states support for LDAP, Active Directory, SCIM, SIEM streaming, KMS and HashiCorp Vault.Read 14 Oct 2026 · datastrato.ai
Semantic layer: An Open Semantic Interchange registry lets teams define metrics once and serve governed definitions to BI tools, notebooks and agents.Read 14 Oct 2026 · datastrato.ai
System requirements: Published minimums are Kubernetes 1.29, Helm 3.8, two CPU cores, four GiB memory and a default storage class.Read 14 Oct 2026 · datastrato.ai
Table maintenance: Per-table policies support compaction, snapshot expiration and orphan-file cleanup through a pluggable JobExecutor SPI.Read 14 Oct 2026 · datastrato.ai

Plans and prices

Datastrato Enterprise Priced on request commercial Apache Gravitino distribution · Kubernetes Helm deployment · license key required · trial keys available · supported release line Read 14 Oct 2026 · datastrato.ai

Every Datastrato Iceberg Lakehouse Platform plan and price →

Strong points and limits

Strong points

Federated registration without metadata copying
Standards-compliant Iceberg REST Catalog
Role, row, column, and tag controls
Scoped credentials for major cloud storage
Maintenance jobs dispatch to several platforms

Limits

Deployment requires Kubernetes and a license key
Security documentation requires NDA access
Commercial distribution is paid

Who it suits

Datastrato targets enterprise teams and developers working with metadata management, governance, and AI-ready infrastructure. It suits environments coordinating lakehouse catalogs, multiple query engines, and controlled agent access.

Verdict

Datastrato provides a federated Iceberg catalog with broad source registration, governance, agent access, and operational maintenance. Its Enterprise deployment is Kubernetes-based, commercially licensed, and priced on request.