Users have access to a number of capabilities that are pre-trained to respond when users ask specific questions. They can select from a number of different response types to configure the way that their agent will respond, and, if desired, they can direct their customers to a live agent to handle more difficult questions.
Create New Agent
When users create a new agent, they can select the type of agent that is most suited to their business needs. Each agent type comes with a specific set of pre-trained capabilities, selected specifically for that industry.
Testing the Agent
As users configure and update their agent, they are able to open a test chat panel and interact with their agent at anytime. This allows them to better see and understand how the agent is responding based on their configuration selections.
Extending the Agent
Users can further extend their agent by connecting to a workspace in the Watson Conversation Service. This allows them to create their own additional capabilities, which can be helpful for adding more unique capabilities for their use case.
Watson Virtual Agent
2-6 designers, 5 offering managers, 10+ developers
IBM Watson brought together a team to create a SaaS, consumer-grade solution for customer service, leveraging the existing machine learning technology.
Our challenge was to create an experience allowing Line of Business users to create a chatbot that is meaningful to their business, without the need for deep technical knowledge. With a combination of pre-built, pre-trained content and the ability to extend their agent’s capabilities with the Watson Conversation Service, users can create an agent that works out of the box, and continue to refine and improve their agent over time.