How graph.one works

Private, social, and public signals are translated into shared graph primitives that improve multiple workflows together.

Graph.one strengthens the existing stack rather than trying to replace every system of record.

View API Docs
Input

Signals that compose

Three classes of signal compose into one usable graph layer.

Private signals

Email metadata, calendar events, CRM records, notes

Interaction evidence from your team's own systems. Graph.one reads metadata only — never email or meeting content.

Social signals

LinkedIn connections, X followers, mutual contacts

Declared and observed relationships across social platforms, imported via data exports.

Public signals

Company data, role history, public profiles

Publicly available information used to enrich and verify identity and affiliation.

The resolution pipeline

Resolution happens before any workflow runs. That's why search, routing, and memory all improve together.
1
Ingest

Connect sources

Email, calendar, CRM, social exports, and public sources feed into the resolution pipeline. OAuth for email/calendar, file upload for social exports.

2
Resolve

Merge into entities

Identity resolution merges duplicate records, tracks role changes, and creates one stable UUID per real-world person or organization.

3
Enrich

Attach relationships and context

Connection strength, interaction history, and team context are pre-computed on the resolved graph before any query or workflow runs.

Output

What the graph produces

The resolved graph exposes these primitives to every workflow.

Stable identity

One UUID per real-world person or organization, surviving job changes, email switches, and alias variations.

Relationship strength

Evidence-backed connection scoring computed from private, social, and public signals. Not a single opaque number.

Network paths

Ranked routes between people through the team's network, with per-leg strength and explainable evidence.

Accumulated context

Meetings, notes, interactions, and role changes attached to the stable entity over time.

See the graph layer in action