There are different options for implementing the digital assistant solution. While a digital assistant solution by nature may imply the use of a human-like visual interface, in practice there are valid reasons to consider different implementations in some scenarios.

The two different options for deploying a digital assistant solution include:

• “Full” digital assistant—As the name suggests, the full solution allows the implementation of a digital assistant in support of its three key modalities: avatar, voice, and text chat. Each end user can turn on or turn off those options, as well as closed captioning. The full option uses all four core components of the solution: 

• Digital Human Platform
• Orchestration and Conversation Management
• Information Retrieval
• Large Language Model

• “Headless” digital assistant—Although not necessarily intuitive, there are cases when a digital assistant solution may be deployed without the 2D/3D human-like interface and the Digital Human Platform. 

This scenario might be required either as a primary or an on-demand approach, with a limited set of modalities, notably as a text chatbot with the voice option. For example, the compute and storage in the environment may be capacity-constrained (for example, under heavier use than planned), or it has been dynamically determined that the bandwidth of the network through which an end user connects to the solution is insufficient for the full experience. 

Under these circumstances, the enterprise providing the solution might not offer the digital assistant avatar frontend, and offer a reduced, traditional chatbot and voice experience without compromising the quality of the responses. Yet the same entire knowledge base is still available.

The choice of which approach depends on the intended use cases of the organization.