One account graph across the systems you already run.
SixDegree resolves entities, relationships, and source observations across the systems your business already runs. Revenue teams, additional applications, and authorized agents use the same operational graph.
Start with forecast, renewals, and deal risk. Expand without building another disconnected data project.
The graph is derived from the systems already running your business.
SixDegree reads the operational systems you already use and turns their records and events into observations about the business. Nothing is migrated and nothing is rewritten. The source systems stay the sources.
Observations, not verdicts
SalesforceOpportunity$1.2M ARR, commit
StripeInvoice$1.05M recurring
ZendeskTicketP1 escalation
JiraIssueDependency delayed
Records become entities when SixDegree knows what they describe.
The same customer, person, product, opportunity, or issue appears differently in every system that touches it. SixDegree resolves those records around the operational entity they describe.
Accounts are the easy case. The same resolution runs across every kind of thing your systems describe, which is what separates an operational graph from a customer view.
- Account
- Person
- Opportunity
- Product
- Issue
- Invoice
- Project
- Dependency
Identity is useful. Relationships make it operational.
SixDegree continuously derives how entities relate across systems, so an event in one part of the business can be understood in terms of everything it affects.
- Personstart
- works_atAccount
- ownsOpportunity
- depends_onCapability
- blocked_byIssue
- impactsRenewal
An engineering dependency slips. The graph resolves which capability it blocks, which opportunities depend on that capability, who owns them, and how much of the forecast is exposed.
Different systems can be right about different things.
SixDegree preserves each observation, where it came from, and when it was observed. Nothing is overwritten to manufacture agreement, because the gap between two accurate systems is usually the thing worth knowing.
Disagreement is information.
Salesforce
Contracted ARR
$1.2M
Stripe
Billed recurring
$1.05M
GainsightHealth
Green
Zendesk
P1 escalations
2 open
What the graph keeps with every claim
$1.05M billed recurring revenue
- Source
- Stripe
- Observed
- 4 min ago
- Entity
- Meridian
$1.2M contracted ARR
- Source
- Salesforce
- Observed
- 2 hours ago
- Entity
- Meridian
One graph. Multiple ways to use it.
- SixDegree
- Ask and explore operational evidence directly, with the trace behind every answer.
- MCP
- Give Claude, ChatGPT, Cursor, and other MCP clients access to the same graph.
- API and SDK
- Build source-backed operational workflows directly on SixDegree, in Go or TypeScript.
SixDegree exposes the operational graph over MCP, so the AI tools your teams already use can reason over the same source-backed evidence.
Read the graph. Act through the systems.
Connected systems expose actions as well as observations. Agents can use those actions where they are explicitly allowed to, and higher-risk writes are gated by approval.
- EvidenceThe graph resolves the account, its owner, the open opportunity, and the support escalation behind it.
- DecisionThe agent reasons over those records and the observations attached to them.
- Scoped actionIt calls one action the connector exposes and the agent is permitted to use.
create_follow_up_taskrecorded against the runContext is shared. Access is not.
- Run as
- Every agent runs as an explicit principal rather than as the platform.
- Scope
- Effective access is that principal’s permissions intersected with the agent’s scope. An agent can only ever do less than the person it runs as.
- Budget
- Per-run ceilings on tool calls, tokens, cost, and time. An agent stops at the budget, every time.
- Audit
- Each run binds to a specific agent version and records the full trace, so you can answer why an agent did what it did.
Integrations expose more than records.
Every connector in the registry declares the same five facets, which is how one connector can contribute to discovery, action, and rendering at once.
- Discovery
- The entities and relationships a connected system exposes, ingested continuously rather than once.
- Actions
- What the connected system allows SixDegree to do, as typed tools an agent can call.
- Visualizations
- How that system’s native objects are rendered, in the web UI or in an AI client.
- Skills
- Reusable multi-step workflows composed across connectors.
- Agents
- Prebuilt agent capabilities, where the connector ships them.
Built to extend.
- MCP
- Use the graph from Claude, ChatGPT, Cursor, and other MCP clients.
- SDK and API
- Build directly against SixDegree, or ship your own connector in Go or TypeScript.
- Models
- Run on Claude, GPT, Grok, or open models through Ollama. SixDegree grounds whichever you deploy.
The same graph supports revenue, customer, support, engineering, and operational work, because those workflows already share entities and relationships. Most teams start with the Revenue Graph.
See what your systems already know together.
Connect the systems your business already runs and see the entities, relationships, and observations SixDegree resolves between them.