Production-ready · Open deployment · AI-native

Stop cataloging data.
Start governing it.

The unified platform that connects every data asset, enforces governance automatically, and puts an AI agent in the hands of every team — from engineers to executives.

terminal
$ make dev
# PostgreSQL, Kafka, Auth, API, UI — all running.
$ open http://localhost:8080
Connects with the platforms you already use

Sound familiar?

A data engineer needs access to a customer dataset for a critical model. She searches three different catalogs, finds conflicting metadata, sends an email to someone who might be the data owner, and waits. Five days later — after a Slack thread, two approval emails, and a Jira ticket — she gets read-only access to the wrong table.

Meanwhile, the governance team is tracking 2,000 contracts in spreadsheets. A schema change broke four downstream dashboards last week — nobody saw it coming because lineage lives in someone's head. The CISO is asking for a compliance audit, and the team is pulling data from six different tools to assemble a report that's already outdated.

Meshport eliminates every one of these problems. In one platform.

Before & After Meshport

Concrete improvements, measured in real workflows.

1-3 weeks
< 2 min
Data access request time
2-4 hours
< 30 sec
Data discovery time
2-3 weeks
Always ready
Audit preparation
5-8 tools
1
Governance tools needed

See the difference for yourself

Deploy Meshport in minutes and experience governance that actually accelerates your team.

Book a Demo

One platform, not ten

Meshport replaces a fragmented stack of tools that were never designed to work together.

📚
Collibra / Alation / Atlan
Unified asset catalog with contracts & lineage
📄
Confluence / Notion
Built-in wiki with collaborative editing
🎫
Jira / ServiceNow
Self-service marketplace with approval workflows
📊
Spreadsheets
Schema-first data contracts with lifecycle management
💬
Slack threads for governance
Change requests with audit trail & threaded comments
🔬
Custom lineage scripts
Interactive lineage graph with impact analysis
📈
Multiple dashboard tools
Role-based analytics with 15 customizable widgets
🤖
ChatGPT / internal bots
73-tool AI agent that executes governance tasks

Everything you need to govern data at scale

Six integrated modules that cover the full data governance lifecycle — from contract to consumption.

📜

Contract Management

Define, version, and enforce contracts with schema validation, quality rules, and a 7-stage lifecycle from draft to retirement.

Schema validation Quality rules Version history Diff viewer
🏪

Data Marketplace

An internal app store for data products — with ratings, SLAs, sample data, certification tiers, and one-click access requests.

Certification tiers Reviews Sample data SLA tracking

Governance Engine

Visual rule builder, change request workflows, multi-level approvals, and an immutable audit trail for every action.

Graph rule builder Approvals Audit trail 6 RBAC roles
🤖

AI Agent

A 73-tool AI agent that can actually do governance work — create contracts, approve requests, run impact analysis, generate reports.

73 tools Multi-turn 7 providers Air-gapped
🔀

Data Lineage

Interactive force-directed graph showing upstream/downstream dependencies, domain clustering, and change impact analysis.

Visual graph Impact analysis Domain clustering
📝

Documentation Wiki

Built-in knowledge base with rich editing, real-time collaboration via Yjs, version history, templates, and code blocks.

Yjs collab Versioning Templates Code blocks
meshport.local/dashboard
Contracts
142
+12 this week
Data Products
87
6 pending review
Active Users
324
48 online now
Compliance
94% compliant
Contract Activity (30d)

Replace your fragmented stack today

One platform for contracts, products, lineage, access, docs, and AI. No more tool sprawl.

Book a Demo

Not a chatbot. A 73-tool AI agent.

Competitors bolted a chatbot onto their search bar. We built an agent that can actually do governance work — create contracts, approve requests, run impact analysis, generate reports.

Think Claude Code, but for your data platform.

Multi-turn conversations. Tool execution with full visibility. Undo any action. Four operating modes for different trust levels.

💬
Normal Mode
Agent asks before making changes. Best for day-to-day work.
📋
Plan Mode
Agent creates a step-by-step plan for review first. Best for complex operations.
Auto-Accept Mode
Agent executes autonomously. Best for trusted, repetitive workflows.
🔍
Discover Mode
Read-only exploration with limited tool access. Best for onboarding and discovery.
73 tools Multi-turn Full undo Risk indicators Slash commands Export Token tracking
M
Meshport Agent
Online
Show me active contracts without quality rules
search_contracts(status: "ACTIVE", has_quality_rules: false)
Found 12 active contracts without quality rules:
ContractOwnerUpdated
customer_eventsTeam Alpha2 days ago
order_transactionsTeam Beta1 week ago
product_catalogTeam Gamma3 days ago
Add a completeness rule to all of them
I'll add a completeness rule to all 12 contracts. This creates 12 quality rules. Shall I proceed?

Runs on any AI provider

Cloud, self-hosted, or air-gapped. Switch providers at runtime — no redeployment needed.

Anthropic Claude
OpenAI GPT-4
Azure OpenAI
Ollama (self-hosted)
AWS Bedrock
Google Vertex AI
Mistral
🛡

Air-gapped AI for regulated industries

Run Meshport entirely on-premises with Ollama. Your data, your models, your network. Zero external API calls. Full audit trail. Complete data sovereignty. Finance, healthcare, defense, semiconductor — this works today, not "coming soon."

Full-stack observability, built in

Monitor every layer of your data governance platform with production-grade tooling — no assembly required.

📈

Prometheus Metrics

Application and infrastructure metrics scraped automatically. Request latency, error rates, queue depth, and custom business metrics.

Auto-scrape Custom metrics Alerting
📊

Grafana Dashboards

6 pre-built dashboards covering API performance, Kafka throughput, database health, WebSocket connections, AI agent usage, and infrastructure overview.

6 dashboards Pre-provisioned Real-time
📋

Loki Log Aggregation

Centralized structured logging from every service. Query, filter, and correlate logs across API, worker, and infrastructure containers.

Structured JSON Log correlation Promtail
🐛

GlitchTip Error Tracking

Self-hosted error tracking and performance monitoring. Sentry-compatible SDK, automatic error grouping, and release tracking.

Self-hosted Sentry-compatible Auto-grouping
0+
API Endpoints
0+
AI Agent Tools
0+
DB Migrations
0
AI Providers
0
Operating Modes
0
Grafana Dashboards

How Meshport compares

Feature-by-feature against the tools enterprises are evaluating today.

Capability Meshport Collibra Atlan DataHub Alation
Data contract management Full lifecycle Limited Basic Plugin No
AI agent with tool execution 73 tools, multi-turn Search chatbot Search chatbot No Search chatbot
Self-hosted / air-gapped Yes SaaS only SaaS only Yes SaaS only
Built-in documentation wiki Full wiki + Yjs collab No Basic No No
Change request workflows Built-in, multi-approver Separate module No No No
Data product marketplace Full (ratings, SLAs) Basic catalog Basic catalog Basic catalog Basic catalog
Real-time collaboration WebSocket-native No Limited No No
White-labeling Full No No No No
Pricing model Transparent $$$$ $$$ Free (limited) $$$$

Ready to leave the competition behind?

Deploy the full platform on your infrastructure. No SaaS lock-in, no per-seat surprises.

Book a Demo

Connects to everything you already use

Sync metadata bidirectionally. Authenticate with your existing identity provider. Stream events to your infrastructure.

Databricks Unity Catalog

Catalogs, schemas, tables, column metadata

Production

Collibra

Lineage, quality scores, glossary, classifications

Available

Microsoft Purview

Lineage, classifications, glossary, ownership

Available

Atlan

Lineage, quality, glossary, usage stats, sample data

Available
🔒

Azure AD / OIDC / SAML

Multi-SSO, auto-provisioning, role mapping

Production

Kafka / Event Hubs

11+ event topics, at-least-once delivery

Production
🤖

7 AI / LLM Providers

Claude, GPT-4, Ollama, Bedrock, Vertex, Mistral, Azure

Production
🔧

Your Own Connector

Implement one Go interface. No core changes needed.

Pluggable

Built for the people who actually use it

Every role gets the experience they need — without learning a new tool for each workflow.

💻

Engineers

Data & Platform Engineers
  • 150+ REST API endpoints — automate everything
  • Pluggable connector architecture (one Go interface)
  • Subscribe to Kafka events from your own services
  • Interactive API playground and developer portal

Data Stewards

Governance & Quality
  • AI agent handles routine tagging, classification, checks
  • Approval queue with threaded discussions
  • Visual governance rule builder
  • Immutable audit trail for every action
🔍

Data Consumers

Analysts & Scientists
  • Browse the marketplace like an app store
  • Preview sample data before requesting access
  • One-click access requests (minutes, not weeks)
  • AI assistant answers "where is the data for X?"

Platform Teams

DevOps & Infrastructure
  • Helm chart with prod-ready defaults and scaling
  • Full observability stack (Prometheus, Grafana, Loki)
  • Health probes, graceful shutdown, PgBouncer pooling
  • Air-gapped deployment for regulated environments
🔒

CISOs

Security & Compliance
  • 6-role RBAC with feature-level permissions
  • Security classification enforced on every asset
  • Session management with individual revocation
  • Air-gapped deployment with zero data egress
📊

CDOs & CTOs

Data & Technology Leadership
  • Domain health scores and compliance metrics
  • Consolidate 5-8 tools into one platform
  • Predictable costs — no per-seat SaaS scaling
  • Real-time governance visibility, not quarterly reports

Ready to unify your data governance?

From zero to a fully running platform in under two minutes. No credit card, no sales call, no vendor lock-in.

Quick start
$ git clone https://github.com/meshport/meshport && cd meshport
$ make dev
$ open http://localhost:8080
# That's it. You're running.
Book a Demo →