User Context Service

One core capability of the Conversation Platform is the power to use data/knowledge to intelligently drive user interactions towards self-service, leading to better and personalized user experiences. Below are the main categories of data/knowledge which are leveraged to improve various Virtual Agent skills, agent tools, and other CCaaS applications.

Conversational data

• Data generated as users interact with the platform is important for driving many of the core capabilities in self-service workflows and agent-handled scenarios.

  • This data also helps to categorize users based on various factors such as agent performance, ease of intent fulfillment,and sentiment computation.
• Additionally, data from the platform can be fed into machine learning modules (connected via ICD) to predict and suggest possible next steps to Virtual Agents. This eventually enhances the user experience while reducing interaction time.

Non-conversational knowledge

• Data associated with partner users is not present on the platform, yet it is crucial in providing contextual insights.

  • The idea of User Context is to define patterns for integrating this off-platform partner data like Loyalty Details, user bookings, user preferences etc. to open up powerful data driven decisions.

Capabilities

• A well-defined API that exposes key user/booking attributes.
• Synergistic integration of multiple data services/sources to enable consumers to use meaningful information.
• High-speed, low latency service with data retrieval optimisations.
• Transformed,Derived and Enriched data exposed for each partner leveraging centralised access management.

Architecture

The following diagram shows the various components of the CCaaS system and where the User Context Service (UCS) fits in.

Users

Users are partner users, travelers, suppliers interacting with Platform applications.

Platform components

Various platform components come into the picture when users interact at both design time as well as run time through the Virtual Agent Control, VNext, and Console.

Partner services

The User Context Service uses partners services and systems to gather user details and data at run time and to generate data for the platform.

Data transformations & enrichment

This layer of the User Context Service consists of the core logic to:

1. Transform, enrich, and unify the partner data into a single schema adhering to platform grammar.
2. Optimize various run time data fetches to be high speed low latency service.
3. Apply partner defined, extensible context definitions to provide partners the ability to define or override context.

API Interface

The User Context Service exposes a GraphQL API to fetch various user and booking attributes so that consumers can decide what data they need from the entire data graph schem

Consumers

Platform components like Intelligent Services (ICD), Agent Tools (VNext, Soft-phone) and VA Skills consume this contextual data to make better and personalized decisions.

Context schema categories

Grain	Category	Description
User	Demographics	User demographics like firstName, lastName, and address.
User	Identifiers	Various user identifiers captured on Partner systems like expUserId, emailAddress, and phoneNumber.
User	Loyalty details	Key details around traveler's loyalty programs.
User	Bookings	The last three years of a traveler's bookings filterable by tripStage, status, or other data.
User	Conversations	The traveler's relevant past conversations to provide historical contextrsations to provide historical context.
Booking	Attributes	Booking grain attributes like trip dates or status.
Booking	Itinerary context	PPartner and business segment details of the booking along with financial catrgories like MU or companyCode.
Booking	User details	Attributes of the user associated with the booking.

Note: Loyalty details, bookings and conversations are clubbed inside travelerContext in the User Context Schema. UCS provides context for any type of user (traveler, agent, supplier) and each can have a different context so that the schema can evolve naturally.

For more information about the User Context Schema, please refer to the schema details.

Getting started

The User Context Service tutorial explains how the User Context Service can be integrated with various services and components for data enrichment or retrieval.