See How Our NLU & ML Based System Makes Your Agents More Efficient
Let’s start with the basics. BotCentral chat is available to run as an AI-driven chatbot within your website, your iOS or Android app or in popular messaging platforms like Facebook Messenger.
Beginning with a user's chat input or question, the BotCentral platform uses our Natural Language Understanding (NLU) technology to match against a set of entities and intents that the chatbot understands. This includes the "seed" set of intents as well as a "learned" set created through user feedback and machine learning. For example, “what is your normal delivery schedule” could be in the seed set; over time, the bot learns that, “how long does it take to ship” maps to the same intent.
After ascertaining intent, the platform will determines whether or not any matched intents correspond to the available Knowledge Base entries, if so, they are returned as conversational messages.
If matched, any Enterprise APIs or other REST APIs are called and their results returned. For example, “when does your store in Hoboken, NJ open on Sundays”, would cause a Store Hours API to be called for the given location in order to provide the right answer.
The chatbot may not know the answer, or the user might want to talk to to a real person. If a user has requested help from an Agent or if a live Agent supervisor determines it’s time to step in and help, the live Agent can join the conversation for one on one communication with the user.
The chatbot conversation with the BotCentral platform or with a live Agent continues seamlessly from your user’s perspective. The chatbot changes its icon to a person when talking to a live agent in order to ensure full transparency and to avoid end user confusion.
The final message is generated and returned to the user. The final message may not be a single message but may be a full follow-on conversation with clarifying questions and additional information. The goal is to solve the problem or issue raised by the user’s original question or input to their full satisfaction.
The BotCentral platform integrates automatically with enterprise platforms such as Zendesk and Salesforce Desk. Automatic integration means that the chatbot learns from and stays-up-date with the enterprise Knowledge Base and logs. it also means that the chatbot-to-human handoff occurs directly into the existing ticketing system used by Agents. BotCentral additionally integrates into the appropriate Support channel(s) in Slack, for teams that leverage Slack for real-time notifications.
Keeping your Customer Support team up to date with the latest Knowledge Base entries and issue resolution learnings is an ongoing problem that requires a scalable solution. Keeping your Knowledge Base up-to-date with what your user’s are really asking about is also a critical problem.
With our BotCentral Advisor Dashboard, we use the latest Machine Learning techniques, enabling your team to automatically suggest the best possible answers and actions available for each issue, and make it easy to share with those answers with your customers. Achieve higher satisfaction from your customers and your Support team.
The BotCentral Advisor is also available as a browser widget and/or Zendesk app - making it conveniently available within or alongside the main issue resolution interface that your Agents are familiar with.
Training new Support staff is a similar challenge. The BotCentral Advisor Dashboard can also be used to help on-board new team members using the same Machine Learning capability.
You can analyze the coverage of your Knowledge Base at any time by bulk loading a snapshot of your existing tickets or chat logs. The BotCentral platform will provide an intelligent summary of hits and misses in your existing Knowledge Base from historical queries. The summary indicates prioritizes the most frequent misses by question, intent, tags and keywords so you can update your Knowledge Base in order to best answer the most frequently missed questions.
We don’t replace your existing systems, we make them work better.