Provide a consistent chatbot experience by linking individual (standard) bots together to create a Universal bot – allowing enterprise users to access the capabilities of multiple chatbots from a single chat and bot persona.
Multifunctional, Multitasking Bots
The traditional enterprise approach to bot building is for each function or department to develop their own specialized chatbots. With increased chatbot usage, however, more and more chatbots will need to be made available to enterprise users. This approach means that users have to interact with multiple chatbots, remember the specific capabilities of each, and have to adjust to the different conversation styles required. A Universal bot bot eliminates the need for users to have to talk to multiple bots to get things done, increasing customer satisfaction and improving the overall productivity of employees.
With the Kore.ai Bots Platform, developers can create a Universal bot and link individual (standard) bots to it. Universal bot provides a consistent experience to users in terms of discovering information and performing tasks. Universal bot can even be enhanced to provide a rich messaging experience to users and – since standard bot responses are delivered through the Universal bot – they too can leverage the rich messaging experience.
Building a Universal Bot
How Universal Bots Work
Universal bots take user requests and determine which linked bots to call to tackle the task at hand, quickly multitasking on the person’s behalf while continuing the existing conversation.
- Acts as a single, unified interface for bot users.
- Allows bots to be built independently for specific use cases and then linked together to form a Universal bot.
- Performs intent recognition across all linked bots to understand the user’s intent, ranks them based on relevance, and engages the appropriate bot to perform the task.
- Handles ambiguous, unrecognized requests by providing the user with matching bot names, tasks, or suggestions so they can quickly make the intended selection.
- Can be deployed to the same or different communication channel as its linked bots.
- Allows developers to view and analyze all user interactions with the Universal bot and test them in batches to discover single matched intents, multiple matches with the same score, and unmatched intents. It also allows developers to train them to match with the correct linked bot and the correct task within each bot. When a Universal bot is trained, all linked bots can be automatically trained as well.
Implement Universal Bots to Accelerate Your Chatbot Strategy
Universal bots are architected to minimize maintenance requirements for developers. Linked bots can be quickly
and easily added or removed without impacting the Universal bot or altering linked tasks and functions.
Use Cases
Banking Bot
Banks offering various products and services to their customers can independently develop product specific bots and incrementally launch them, adding them to a common Universal Bot.
Banks offer a wide range of financial services and products, such as credit cards, savings accounts, and business loans. Universal bots allow banks to develop product or service-specific bots and incrementally launch them when ready – without impacting bots based on other products – by adding them to a single, unified bot for customers to interact with.
Release cycles of these product or service-specific bots can be maintained independently so that a specific group can upgrade their bot without waiting for other groups within the organization.
For example, a financial institution can develop one bot to answer questions related to credit cards and another later to answer questions about business or personal loans. Rather than force customers to reach out to multiple bots, these bots can be integrated into a single Universal bot. When a customer needs support related to both of these topics, a Universal bot is evoked that then talks with both bots and sends back the required information to the user.
HR Bot
HR bots can be built incrementally by integrating various departmental bots as and when they are ready. A Universal bot could, for example, be launched with one or two standard bots that answer common questions related to HR policies. This could then be gradually scaled up by integrating other bots that perform more advanced functions like leave management, performance reviews, and more. End users could thus perform new tasks, including those within different back-end systems, without ever needing to switch between multiple HR bots.
