Experience Memory
Experience Memory brings together signals from across your business, including every channel
Various voice and digital communication mediums that facilitate customer interactions in a contact center., every system, and every customer touchpoint, into a single, continuously updated view of each customer. It makes that view available in real time to every application in your stack, so routing, AI agents, copilots, campaigns, and your existing business systems can all act on the same current, complete picture of the customer.
Signals come from many places: web and mobile activity, brick-and-mortar visits, business systems such as CRM Third-party systems that manage such things as contacts, sales information, support details, and case histories., ERP Enterprise Resource Planning. Software system that manages and automates core business processes., billing, connected devices, and every customer interaction. Most organizations already capture these signals for operational purposes, but they rarely reach out to the applications that shape the Customer Experience. Experience Memory closes that gap without replacing the systems you already have in place.
Experience Memory is made up of four components, each with a distinct role:
| Flow | Details |
|---|---|
|
Integration Fabric |
Connects every signal source, regardless of protocol, format, or data stream, into a single, unified feed, and streams enriched data back out to your existing business systems. |
|
Active Memory |
Maintains a real-time, identity-resolved profile of each customer built using multiple real-time signals. This also includes Common Data Objects and System Data Objects. |
| Weaver | Analyzes signals and customer journeys to identify patterns, friction points, and escalation triggers. |
| Friction Intelligence | Turns signals and Weaver-composed journey narratives into deterministic, auditable actions: routing, notification, automation, or escalation. Also known as Signal Suite. |
| Signal Engine | Action and automation layer. Evaluates configurable rules against live journey signals and inbound data, dispatches connector actions and notifications, and orchestrates multi-step workflows. |
Integration Fabric
Integration Fabric is the inbound and outbound data layer for Experience Memory. It connects your signal sources, regardless of protocol, format, or data stream, into a single, unified feed, and resolves differences between them automatically, so every downstream component works from clean, consistent data. It also streams enriched signal data back to your existing business systems in real time, so the rest of your stack always has access to the same current customer picture that Experience Memory holds.
Integration Fabric handles:
- Connector management for external and internal signal sources.
- Credential management for connected systems.
- Intelligent protocol management, translating between formats automatically.
- Outbound data streaming to downstream systems, including CRM Third-party systems that manage such things as contacts, sales information, support details, and case histories., analytics platforms, and custom business applications.
Because Integration Fabric absorbs this complexity, you don’t need to rebuild your existing data architecture to bring a new signal source online.
Active Memory
Active Memory is the identity and profile layer of Experience Memory. It performs identity resolution in real time, stitching new signals and contact events into a comprehensive customer profile the moment they arrive.
Active Memory is not a historical archive: it is a picture of the customer as of the latest signal, always current. It is built from:
- Identity and attributes: the core profile data associated with a customer.
- Identity resolution: a single source of truth for customer identity across the platform.
Common Data Objects (CDO) and System Data Objects (SDO)
Common Data Objects and System Data Objects both live inside Active Memory and share the same underlying schema and storage infrastructure. The boundary between them is architectural: it is determined by which side of the platform boundary the creating application sits on, and enforced at object creation time by the authentication token the application presents.
| Common Data Object | System Data Object | |
|---|---|---|
| Created by | Your customer or partner applications, authenticated with a GA token, through a configuration UI or directly via API. | NiCE applications and platform services, authenticated with an internal service token, at runtime. |
| Purpose | Model your data exactly the way you want to view and use it, usable by every CXone application and every other application in your stack. | Hold high-volume, operational data that the platform needs to run. |
| Visibility | Viewable in the platform UI or retrieved via API. | API only, not exposed in the platform UI. |
| Schema governance | Customer-defined; versioned and lifecycle-managed. Purging data does not delete the schema. | Application-defined; managed by the owning service. Purging data does not delete the schema. |
| Lifecycle and deletion | Business-rule-driven; deletion is triggered by schedule, event, or customer decision. | Owning application signals deletion explicitly on interaction or task completion; data is retained until that signal is received. |
Both object types hold operationally relevant, real-time data rather than acting as a long-term historical archive. Both are tenant-isolated, support sub-3-second retrieval, and separate schema from data: purging data instances does not remove the schema definition.
Weaver
Weaver is the intelligence layer that sits between Active Memory and Signal Engine. It reads the longitudinal context held in Active Memory and analyzes customer actions across sessions and channels, weaving individual signals into coherent journey narratives that Signal Engine can act on deterministically.
Weaver identifies:
- Friction points in the customer journey.
- Behavioral patterns.
- Escalation triggers.
Rather than reacting to a single event in isolation, Weaver looks at what a sequence of events means, and generates the narratives and classifications that drive the action taken next.
Signal Engine
Signal Engine turns signals into action. It translates raw signals, or the journeys and intents that Weaver composes, into deterministic, auditable actions: routing, notification, automation, or escalation.
Because Signal Engine actions are deterministic and auditable by design, every action it takes can be traced back to the signal or narrative that triggered it.
How It Fits Together
Experience Memory is not a system you log into on its own: it is the layer your entire stack draws from:
- Signal sources flow in through Integration Fabric, which normalizes and unifies them before passing them to Active Memory.
- Active Memory resolves identity in real time and maintains the always-current customer profile.
- Weaver reads that profile continuously, identifying patterns and composing journey narratives.
- Signal Engine translates those narratives into routing, notification, automation, or escalation deterministic actions, that are delivered to the applications that need them.
- Integration Fabric simultaneously streams the enriched profile back to your existing business systems, so your CRM Third-party systems that manage such things as contacts, sales information, support details, and case histories., analytics platform, and other applications are always working from the same data.
Together, these four components ensure every AI agent, human agent, and business application in your stack is working from the same complete, real-time view of your customer.