It’s not unusual for companies to have several chatbots. For example, a HR, IT and legal bot for internal support, or a pre- and after-sales one for customer service. To improve the user experience, dydu decided to deploy a metabot feature, enabling users to interact with several bots at the same time via a single interface.
What Is a Metabot?
Metabot is a feature that enables users to call on several bots at the same time through a single chatbox. It then determines the best answer for the end-user. It takes on the form of a parent bot, to which several child bots are attached.
A metabot provides a single entry point for users, whatever their query. The knowledge managers, on the other hand, are only responsible for their own bot(s). The feature is deployed by activating the metabot option in the dydu solution.
Why Implement a Metabot?
Improve Your User Experience
A metabot prevents customers, employees and users wasting their time looking for the right chatbot. They only need to interact with one bot, whatever their question may be. The risk of having several bots without a metabot is that the user interacts with the wrong one and gets frustrated when it doesn’t understand their query.
Take for example an organisation with two internal chatbots, one for HR and one for IT. An employee notices that the number of remaining leave days in the HRIS is wrong. They’re not sure whether to talk to HR or helpdesk. With a metabot, it’s easy: they can simply ask the chatbot! Everyone saves time and gains in productivity.
Improve Access Rights Management
Each child bot has its own environment and knowledge manager (usually an operational employee). The bots’ knowledge bases are therefore managed separately, without impacting each other. Managers can only perform actions on the bot(s) they have access rights to.
Each bot also has its own dialogues. Knowledge managers can continually improve their bots (reading and correcting conversations and misunderstood sentences, audits and suggestions) without worrying about the others.
A metabot manager oversees the management of all the child bots.
A Global and Specific Chatbot View
Metabot managers need to be able to monitor and analyse performance without accessing each child bot. They have an overall view of the parent and child bots in terms of analytics and learning modules. This enables them to help knowledge managers achieve greater consistency and to arbitrate when two subjects cross over.
Dydu deployed the first version of its metabot in 2015. We have made several improvements to our new version in order to enrich the end-user experience for multi-bot companies and make bot management even easier.