The intelligence layer for
your GTM relationships.

Know what changed. Understand why.
Give your team the context to make the next move.

Open dashboard

Open source. Evidence first. Built to connect.

Onecanonical history
Everysource, connected
Yourrules, in control
Illustrative restaurant scene: a sunlit stone storefront with teal doors and an owner preparing the dining room.Maison Olive

Second location announced

Official announcement
02 OCT · 10:00 UTC
GTMAG
CONTEXT READY
Evidence attached

Source · timestamp · provenance

Traceable by design
One history. The source stays attached.
A new location. A useful signal.

An official announcement enters the history with its source and observation time.

Illustrative product flow
PostgreSQLMCPNext.jsTypeScriptGitHub

One history for the
whole relationship.

From the first observation to the next conversation.
Connect the dots without losing where they came from.

Explore the product
gtmag / observeJSON
1POST /api/v1/events2 3{4  "type": "account.location_opened",5  "source": {6    "connector": "manual",7    "rootSourceId": "announcement-01"8  },9  "eventTime": "2026-10-01T09:00:00Z",10  "observedAt": "2026-10-02T10:00:00Z",11  "subjectRefs": {12    "accountId": "…"13  },14  "schemaVersion": 115}
01 / CANONICAL EVENTIllustrative preview

One event. All the context.

Immutable observations, with their source and timestamps.

Maison OliveAccount discovered
Account
Alex MartinVerified provider identity
Person
Second locationOfficial announcement
Event

Intelligence, built into your workflow.

A shared foundation for people, tools and agents.
The same context, wherever your team works.

See how it works

Nothing important
gets lost between tools.

People, evidence and history.
A foundation that holds them together.

One canonical history

Bring commercially relevant events into one durable timeline, with their source and timestamps intact.

Explore the design

Know who you’re looking at

Connect people, accounts and locations through deterministic identity resolution and human merge review.

Explore the design

Keep the evidence attached

Trace each claim to its source. Distinguish what is verified, inferred and not observed.

Explore the design

See the signal in the noise

The V0 design turns small interactions into explainable buildup, rather than an opaque lead score.

Explore the design

Keep your constraints in control

Opt-outs, permissions and contact policies are designed to stay deterministic, regardless of signal strength.

Explore the design

Built for your whole stack

A shared backend for the web, HTTP API, workers and MCP. Bring your tools to the same source of truth.

Explore the design

A seamless relationship
layer for your stack.

People, accounts and evidence. Connected by design.

Explore GTMAG
One connected record
Alex MartinOwner · Maison Olive
Account ownerAlex MartinVerified provider identity

Your sources. One relationship.

Connect the account, the person and the conversation. Keep the source behind every detail.

Observation recorded

A useful signal arrives.

An official announcement is attached to Maison Olive.

ObservationEvidence
Nothing is sent automatically.

From a signal to a considered next step.

See what changed, check the evidence, and decide what happens next. Your contact rules stay in control.

Illustrative product previews · not live customer data

Find the signal.
Keep the evidence.

A fact should come with a source. An inference should look like an inference. Keep the explanation close to every decision.

See the contracts
Illustrative restaurant scene: a sunlit stone storefront with teal doors and an owner preparing the dining room.
Maison OliveA new chapter.
A second address.
EvidenceVerified source

Opening a second location.

The announcement supports this claim.
The original source stays one click away.

New locationVerified
Buying intentNot observed

All your context.
One shared history.

One durable backend across every surface.
Keep the history connected as your
workflow grows.

From observation to action● Canonical context
OBSERVE
RESOLVE
EXPLAIN
ACT
An explainable path. Not a black-box score.
01

Intelligence EngineUnderstand what happened and what the evidence supports.

02

Relationship EngineKeep people, goals and relationship context connected.

03

Activation EngineDesign the next move around evidence and deterministic guards.

Your relationships
stay yours.

Built around traceable evidence and explicit control.
Core intelligence should work without a paid AI API.

Deterministic guards

Hard constraints stay hard. An opt-out must never be overridden by a score.

Immutable history

Events preserve what happened and when it was observed. New knowledge creates new records.

Human merge review

Ambiguous identities stay separate. Canonical merges require a human, evidence and a reason.

Time-aware context

Event time and observed time stay distinct, keeping the real-world sequence inspectable.

WebHTTP APIMCPWorkers
Explore the architecture
Bring in the event.

Source-attributed. Versioned.
Idempotent from the start.

Canonical event
Connect the context.

Identity, evidence and goals.
One relationship history.

Shared intelligence
Make the next move.

Explainable recommendations.
Your constraints in control.

Guarded action

Connect once. Understand more.
Move with context.

From the first observation to the next right conversation.
Build on a foundation your whole team can inspect.

Start building

Evidence and control.
Without compromise.

Keep the source. Preserve the uncertainty.
Make the decision trace inspectable.

Read the specification

Permissioned writesApplication services validate request context and permissions before privileged changes.

Provenance by designSources, observations and evidence belong alongside the knowledge they support.

Open, inspectable foundationExplore the code, contracts and architecture. Core behavior is designed to work without an LLM.

A little clarity,
before you start.

What is GTMAG?

GTMAG is an open-source relationship intelligence and GTM control plane. It brings event history, canonical identity, evidence and relationship context together across your existing stack.

Does GTMAG replace my CRM?

GTMAG is designed to work alongside CRM, enrichment and sales engagement tools. It owns the cross-system history and relationship context that connect those tools.

What is available today?

The repository includes the application foundation, PostgreSQL schema, authenticated event ingestion and deterministic identity resolution with merge review. Evidence, relationship policies and operational surfaces are being built in the P0 roadmap. Product previews on this page are illustrative.

Do I need an AI subscription?

The core is designed to run without a paid LLM or AI API. Structured contracts, deterministic guards and traceable evidence remain the foundation.

Can I run GTMAG locally?

Yes. The repository documents Node.js 22.12+, pnpm and PostgreSQL setup, migrations, development commands and automated checks. Follow the getting-started link in the footer.

Will GTMAG send outreach automatically?

V0 is recommendation-first. It does not include autonomous mass outreach or outbound mail delivery. Future external actions must use explicit, authorized execution paths.

Give your team a way to
see the whole relationship.

One history. Clear evidence. Better next moves.

Explore GTMAG
Opensource foundation
Oneshared backend
Zerorequired LLM APIs