Architecture
System of Context
Rox organizes connected revenue data around accounts, people, opportunities, and interactions. The System of Context gives Rox a shared foundation for account research, meeting preparation, outreach,
For example, a CRM record may identify an opportunity and its owner, while emails and meeting transcripts describe the conversations behind it. Connecting those records allows Rox to use them together when responding to a request.
Sources of context
Rox combines information from the sources enabled for your organization.
Source category | Examples of context |
|---|---|
CRM and structured enterprise data | Account attributes, contacts, opportunity stages, ownership, and configured business metrics |
Communication and meeting data | Emails, calendar events, meeting notes, and transcripts from connected sources |
Public information | Company websites, news, job postings, and other information used in research |
User-provided information | Documents and instructions supplied for supported workflows |
The available information depends on the integration, its permissions, and your organization’s configuration. A source category does not imply that every system in that category is supported.
Entities and relationships
Rox organizes data into entities that represent the people, organizations, deals, and interactions involved in revenue work.
Entity | What it represents |
|---|---|
Account or company | An organization your team researches, sells to, or manages |
Person | An individual associated with accounts and interactions |
Opportunity | A potential deal, with details such as stage, value, and close date |
Event | A meeting or other supported event, with participants and associated context |
Relationships connect these entities. A person can be associated with an account, an opportunity can belong to that account, and a meeting can connect participants with the account’s activity.
Communication records and other context can be linked to these entities so that information from different sources can be retrieved together.
Mapping source data
Source systems use different names, fields, and identifiers. Mappings connect those representations to Rox entities and attributes.
For example, a CRM account’s annual revenue field can map to an account attribute in Rox. A source identifier links the Rox entity back to the original record. Relationship mappings describe how records connect, such as an opportunity’s association with an account.
Mapping establishes how source data is represented. It does not by itself determine whether two records describe the same real-world entity.
Resolving identities
Entity resolution connects records that refer to the same company, person, or event.
The relevant signals vary by entity. Company records may use domains and other identifying attributes; person records may use email addresses and other identifiers; event matching can use information about the meeting.
This reduces fragmentation when the same entity appears in multiple sources. It also requires care: similarly named companies, subsidiaries, and people with changing identifiers should not automatically be treated as duplicates.
Resolution within Rox should not be interpreted as a promise to merge or delete records in a connected CRM. Changes to source systems depend on the applicable integration and writeback configuration.
Retrieving context for a task
The agent needs information relevant to the request. Rox can scope retrieval by account, person, time period, and the information being sought, then rank candidate content for relevance.
A meeting-preparation request might draw on recent conversations and the associated opportunity. A research request might use public information about the company alongside configured account data.
The context available also depends on access permissions. Record access and field permissions are separate controls; being able to access an account does not automatically expose every field or related item.
Keeping context current
Connected sources change over time. Rox processes source updates and refreshes the entities, relationships, and derived information used by its features.
Freshness depends on the source connection and the processing required. Source synchronization, entity resolution, research generation, and application updates can occur at different times. There is no single refresh interval that should be assumed for every type of context.
For connection and synchronization details, see the relevant integration guide. For access configuration and troubleshooting, see Admin Guide → Governance.