Skip to main content

AI agents are only as good as
the context they can reach.

Claude, ChatGPT, Gemini, Copilot: your people use them every day, but none of them can reach how your organization actually works: the systems, the data, what it all means, and who owns it. Marmot is the open source context layer that puts all of it within reach of every assistant, agent, and person.

Get StartedLive Demo

Built by engineers who've shipped at HashiCorp, Adidas, Just Eat Takeaway.com and Traefik, and help maintain Kubernetes, Terraform, Redpanda and the Cloud Native Computing Foundation.

Plugins

Iceberg

Populate

Terraform
Pulumi
Populate
Marmot
MarmotContext layer
Discover

AI Agents

Integrations

Why a context layer

When AI doesn't know, it guesses

When someone doesn't know what a column means or where a database lives, they ask a teammate. An agent can't do that. It only knows what it can reach, so the knowledge people carry in their heads needs a place both humans and agents can actually reach.

Hardcoded today
  • Answers live in people's heads, Slack threads and stale wiki pages
  • Schemas pasted into prompts that quietly go out of date
  • Nothing tells the AI what GMV means or which orders table is the real one, so it guesses
  • Every team wires up the same context by hand, separately, over and over
With Marmot
  • Catalog every service, database, dashboard and pipeline once
  • Ownership, definitions and lineage attached to every asset
  • One MCP endpoint every assistant shares, from Claude to Copilot
  • Always live, never a stale copy

You don't move your data into Marmot. You stop hardcoding the context around it. See Marmot for agents

One catalog for people and AI

Catalog every asset once, from services to databases to dashboards, add what it means and who owns it, and make it easy to find for your team in the UI and for agents over MCP.

Discover

One place for agents and humans to find every service, API, queue, topic and database.

Understand

See how a dashboard connects to a table, and what breaks if it changes.

Contextualize

Ownership, business definitions and custom fields that give AI the full picture.

Share

Expose certified context through MCP, the API and the UI.

Answers, not guesses

The questions that used to land in a team's Slack channel get answered on the spot. Marmot's built-in MCP server gives the assistants your people already use answers backed by your actual catalog.

One server to set up, not one per data source. Works with any MCP-compatible client.

Set up MCP

What tables do we have related to customer orders?

discover_data

I found 3 assets matching "customer orders": the orders table in the warehouse, an orders_raw Kafka topic, and a daily_orders_summary view.

What breaks if we rename the order_gmv column?

get_lineage

The daily_orders_summary view and the Revenue Overview dashboard both depend on that column. The table is owned by the Data Platform team, so check with Sarah Chen first.

And what does GMV actually stand for?

lookup_term

GMV is Gross Merchandise Value, the total sales revenue before deductions. That definition comes straight from your business glossary.

Traditional catalog
ElasticsearchSearch
KafkaEvents
FrontendUI
APIBackend
Neo4jGraph
MySQLMetadata
AirflowOrchestration
7+ servicesHours to deploy
Marmot
Marmot
MarmotSingle binary
PostgreSQLSearch, storage & graphs
2 servicesMinutes to deploy

Less infrastructure, same power

Traditional data catalogs need an entire platform team. Marmot needs a database you probably already run.

Quick start

Deploy in under five minutes

Marmot runs as a single binary backed by PostgreSQL - the only dependency you need to start cataloging your data.

Follow the Quick Start guide
~ / marmot
$docker compose up -d
[+] Container postgres-1 Started
[+] Container marmot-1 Started
Marmot is running at http://localhost:8080

Connect to your data sources

Growing ecosystem of plugins

PostgreSQL

MySQL

MongoDB

ClickHouse

BigQuery

Kafka

Airflow

dbt

S3

DuckDB

SQS

OpenAPI

View all plugins

Don't see your data source? Open an issue to request a plugin.

Built to scale

Simple architecture doesn't mean limited. Our largest open source deployment serves 175+ active users, and load tests go well beyond.

Load tested on real infrastructure
500k+
Assets
100+
Concurrent users
<50ms
Avg response time
Read more

Only metadata. Your data stays put.

A context layer needs to know about your assets, not to hold their contents. Marmot is built so the data itself never leaves your systems.

Metadata, not your data

Marmot catalogs schemas, ownership, descriptions, lineage and statistics, encrypted under your key. The rows, messages and payloads inside your systems never enter Marmot.

Deploy it your way

Use managed Cloud, or run the open source build yourself. Run it yourself and even the metadata stays inside your own VPC and under your own controls.

Open source and auditable

MIT licensed and built in the open. Read exactly what Marmot collects, how it connects to a source, and what it stores, line by line.

Talk to us

One context for people and agents.

Giving your employees and agents a context layer is a huge boost in autonomy and productivity. Explore the live demo, deploy it for free, or talk to us when you're ready to scale your context layer.

View Live DemoTalk to Us

Prefer to explore first? Read the docs