Categories:
Aura Context
Aura Context, Aura’s short-term memory, is a key component that allows the enrichment of the responses provided to the users with information from their previous experiences.
Find in the current documents the description of this component, its architecture, components and processes.
Aura Virtual Assistant component
Introduction to Aura Context
In previous Aura Platform releases, Aura experiences had a lack of memory, as every time the user interacted with Aura, the system took it as a first interaction and, once it had finished, Aura forgot it.
In this framework, Aura Global Team has introduced the concept of context, defined as information surrounding a user interaction. Aura Context aims to enhance the Aura-user interaction by providing the system with short-term memory. This capability enables Aura to understand, act and learn based on recent interactions with the user.
Through the storage of key information regarding the user’s previous experiences and environment in a temporary, high-available, low-latency database during a configurable period of time and its further integration in the conversational flow due to the real-time recovery of data with no lateness, Aura is able to better understand the user. This, in turn, allows it to provide more comprehensive responses and adjust its actions based on this information. The stored data is specific to one user and one interaction of that user.
Aura Context is not a critical module for use cases development but can be used optionally for the enrichment of the overall experience.
Aura Context stakeholders
In Aura Context framework there are two main stakeholders, with differentiated roles:
-
Aura Global Team
- Management of the Aura Context data model
- Edition of the Aura Context global data model
- Feeding of the database with data from different Aura components
- Use of Aura Context by global developers
-
Aura Local Team (OBs)
- Mainly consumers of Aura Context data (although they can also carry out certain modifications over the global data model)
- Storage of data in the Aura Context database
- Making queries to the database
ℹ️ The current documents are focused on the use of Aura Context for Aura Local Teams
What is included in Aura Context v1?
From the general definition and scope of Aura Context included in the previous section, the current version of this component, Aura Context v1 constitutes the first approach of the end-to-end product.
Currently, the context generated by Aura is able to retrieve the following information:
- What has Aura understood?
- What is the current status of the channel/device from which the user is interacting?
For this purpose, Aura Context v1 includes the following basic functionalities:
-
Each OB can create its own agent to start working with Aura context in the aura-bot dialogs.
-
Through this agent, the OB can perform the following actions:
- Queries to the context database in order to retrieve key user’s information.
- Creation of custom context fields, if it is needed to store and query information that is not included in the context data model.
- Storage of data in context, for further use in use case development.
Every information included in the context module is stored in Kernel, so datasets are available for analytical purposes or to create intelligent models upon it, as long as you have the right consent to do so.