AI-powered conversations with a personal touch.
The Story
No one wants be stuck talking to some company’s chatbot or have to wade through a depersonalized “do not reply” marketing email that doesn’t answer their real questions about a company or its products. One of the tasks of the Conversation Design Team within the Product team was to build the lifeblood of Conversica’s product—the AI-powered Conversations.
The conversations are what customers actually buy when they subscribe to Conversica. These conversations use AI to field customer requests—but every bit of content, every turn in the conversation logic, is put together by Conversation Designers.
The conversations begin with an introductory message that is one of many possible iterations written by Conversation Designers. Depending on how the lead responds to the initial message, they would then follow the appropriate path in the conversation logic, and each message along that path would also have been written, in its component parts, by Conversation Designers. Some fields are customizable by customers as well; these are input into the system by the Conversation Designers too.
My Role
Conversation Writer I wrote all of the original the content of the conversations I created. This involved writing many possible versions of each phrase, with and without different combinations of customizable fields that could be added to the text. When writing these customizable conversations, it was important to also edit each phrase for both grammar and the usage of the conversation fields, insuring that when a user input a field value such as, for example “our products and services,” the resulting message would make sense in every possible iteration of the surrounding content.
Conversational AI Logic Designer I designed and created conversation logic flows from scratch.
QA Tester I tested every conversation I created to ensure that every stage of the conversation was accurate, felt personalized, made sense at every stage of the conversation, provided the correct alerts to users based on lead responses, had working links, and handled both usual and anomalous lead responses correctly.
Documentation Writer I wrote and edited the documentation for every conversation I created, as well as all surrounding documentation that could reference that conversation. I became the go-to person for documentation on my team and regularly created and executed plans for documentation edits on a large scale.
Product
Conversations are the essential product offered by Conversica. At the end of the day, Conversica’s customers are interacting with the conversations they have bought in skill packages or Revenue Digital Assistant packages, and they want the conversations to make sense, have structure and backup phrases when the user doesn’t fill in a customizable field, and work without a fuss.
The content and conversation logic needs to reel leads into a comfortable, personalized conversation. The alerts need to make sense to the users, who need to be able to easily look through lead responses that need current action. Conversation Design requires a high level of detail because any mistakes in the process can result in confusing errors for customers and their leads.
My conversations are well-written, well-designed, and thoroughly tested. The conversational logic is sound and I have individually checked that each alert sent to the user is correct based on the conversation logic.