Meshport — The Unified Data Governance Platform

Deployment

Choose your deployment method

Three paths to production. All self-hosted, all fully featured.

▶

Docker Compose

Best for development and small teams. One command spins up PostgreSQL, Kafka, Auth, API, and the UI. Great for trying things out or running a proof of concept.

Development Small teams Quick start

⎈

Kubernetes Helm

Production-ready deployment on any Kubernetes cluster. You get horizontal scaling, rolling updates, resource limits, and health checks out of the box. Battle-tested with AKS, EKS, and GKE.

Production Scalable HA

🔒

Air-gapped

For environments with no internet access. Pre-packaged container images, offline Helm charts, and Ollama for AI. Everything stays inside your perimeter with zero external dependencies.

Offline Secure On-premises

Everything You Need

What's included

Every deployment includes the full platform. No feature gates, no upsells, no surprises.

📋

Data Contracts

Manage contracts through their full lifecycle. Schema validation, quality rules, and governance workflows from draft to active.

📦

Data Products

An internal marketplace for data products with input/output ports, SLAs, and lineage tracking.

🤖

AI Agent

73 tools that handle governance, discovery, and automation. Works with Ollama, OpenAI, Azure, and Anthropic.

👥

Teams & Domains

Hierarchical domains with role-based access control. Stewards, consumers, and admins each get the permissions they need.

📈

Observability

Grafana dashboards, Prometheus metrics, and Loki logs. Pre-configured and ready to go.

🔌

Catalog Connectors

Import from Databricks Unity Catalog today. Collibra, Microsoft Purview, and Atlan connectors in the framework.

FAQ

Frequently asked questions

How long does deployment take?

Docker Compose takes under two minutes. Kubernetes with Helm is typically 10 to 15 minutes including namespace setup and secrets. Air-gapped deployments take about 30 minutes the first time because of image loading.

Is Meshport really self-hosted?

Yes, fully. Meshport runs on your infrastructure. Docker Compose for development, Helm for production Kubernetes. You own the deployment, the data, and the network traffic. Nothing phones home.

Does my data ever leave my network?

Never. All processing happens inside your environment. If you use external AI providers (Claude, GPT-4), API calls go directly from your server to the provider. With Ollama, even AI stays fully on-premises.

What kind of support is available?

Three options. Community support through GitHub issues and discussions. Professional support with priority email and a 24-hour response SLA. Enterprise support with a dedicated engineer and custom SLA. Every tier includes full documentation and deployment guides.

Can I migrate from existing governance tools?

Meshport ships with a production-ready Databricks Unity Catalog connector. Collibra, Microsoft Purview, and Atlan connectors are in the framework. The AI agent can also help you bulk-import contracts and products from spreadsheets and documents.

Deploy Meshport in minutes