Interaction Automation System (IAS)

Learn about the IAS and how it automates interactions with end users.

The Interaction Automation System (IAS) predicts and automates effective actions and interactions for end-users. It clusters attributes for both literal and non-literal meanings and employs short-term conversational memory, enhancing the relevance and context of user interactions. It can also leverage any internal LLMs you utilize.

Advantage

IAS offers you an advanced tool that automates the selection of actions, including content and interactions, to facilitate precise action prediction and high personalization, which in turn enhance engagement and user satisfaction.

Use cases

• You can leverage IAS's predictive capabilities for context-aware interactions to enhance user engagement.

• In applications that require nuanced conversational continuity, IAS’s short-term memory feature ensures that dialogues are coherent and contextually relevant.

Key features

IAS has the following key features:

• Retrieval-Augmented Generation (RAG), which allows for accurate, contextually rich, and personalized actions.

• Ability to reach out to an LLM only when the confidence of RAG system is low.

• Asynchronous personalization

• Grammatical layer that enhances and enriches the responses of IAS

We advise our clients to plug their internal LLMs into EDAA™ (through the IAS framework of custom LLM support; for more information, see IAS frameworks for LLMs) so that EDAA™ can obtain contextual knowledge through an already-trained LLM rather than a rogue open-source one.

IAS frameworks for LLMs

The IAS framework for LLMs facilitates the integration of third-party large language models like GPT, providing flexibility in choosing the LLM that works for you.

Note that EDAA™, the underlying technology that powers Affective Computing, itself is not a large language model. It operates on semantic feeding of actions and attributes. However, in cases when these components are missing in the solution, EDAA™ can refer to the connected LLM (if available).

This feature incorporates advanced speech-to-text features and supports interactions in English and Spanish.

Advantage

Affective Computing (powered by Virtue)'s IAS framework offers seamless integration with LLMs of the client's choice. This guarantees that your solution always provides an informed response to end-users, even in instances where specific content attributes are not available.

Use cases

• If you require enhanced conversational capabilities in their applications, you can achieve it by integrating LLMs like GPT with EDAA™, ensuring stable interactions with users.

• For applications involving multi-lingual interactions, you can leverage the IAS framework's support for English and Spanish to broaden their user engagement.