Chatbots are in your life. And you didn’t even know it. Humans will soon develop meaningful relationships with bots. Successful relationships lead to conversations with multiple people. To keep the conversation in sync, you need an Event-Driven Architecture with socket connections to mobile and laptop endpoint business and consumer apps. Group chatbot apps with ChatEngine and Watson Conversation API has enabled businesses to augment their workforce artificially. With the IBM Watson™ Conversation service, you can build a solution that understands natural-language input and uses machine learning to respond to customers in a way that simulates a conversation between humans.
Bots Joining in on Groups Discussions
Bots are built with special purposes. A support robot is trained with helpful FAQs. A sales bot assists in pricing and packaging discussions. A training bot will provide insights in the form of stories. In product evaluation cycles, there are many stakeholders. Synchronizing bot responses with multiple parties is a challenge today. Stakeholders may not be online at the same time. When should stakeholders receive notifications about relevant topics?
Group Chat Use Case Patterns
Conversations bots, like Amazon Lex and Google DialogFlow, typically communicate in a one-on-one settings. Conversation bots built with a request-response REST API can not participate in a group setting. The conversation is in a request/response loop. In a ChatEngine chatroom, humans will interact with each other in realtime. The group typically engages a chatbot. This allows a group to interact with each other including bots. Group and multi-bot conversations is possible only with Event Driven Architecture with socket connections. This is what is offered by PubNub.
With PubNub + Any Conversation API, stakeholders can be notified in realtime and share conversations. When humans and bots subscribe to a live socket connection, the conversation is sharable. Bot queries are shared with the group, allowing meaningful engaged multi-party conversations.
With a strict request/response communication model, the bot is not able to ask follow-in questions continuing the conversation thread. With an EDA, bots reply asynchronously out-of band, taking a mind of their own. A convincing human exchange is achieved when both bots and humans contribute to the continuation of the thread.
Your conversations have entered the new epoch of AI. Are you talking with a bot or a human? Really at a certain level, it doesn’t matter. But if you need to know you can make a reasonable determination. There are trick questions you can ask a bot. These questions will determine if the person you are talking with is actually a bot. In a few more years these questions will no longer be revealing.