‘Ask for Help’: can a bot receive assistance from a human?


Most probably you have heard about the countless cases where Artificial Intelligence can enhance the operation inside a Contact Center. In many cases, they replace human agents and take calls as voice bots, messaging chatbots, or create hybrid ecosystems, where both bots and humans can interact. But what happens when the bot doesn’t have the answer in its knowledge base or doesn’t have access to external databases? The answer is ‘Ask for Help’, an innovative solution proposed by inConcert.

A bot that asks for help if it doesn’t have the answer

There may be cases where a bot will encounter limitations in its interactions with users and will not have a response:

  • Environments where most of the information of an operation or business is not fully documented
  • Specialized know-how that only employees with years of experience have and have not passed on to others
  • Information so new and changing that it is still in the process of being uploaded to the bot’s knowledge base
  • Information too old to have been considered for addition to the knowledge base

Before completely transferring the interaction to a human agent, inConcert offers a unique solution on the market: Ask for Help. With this tool,bots can request help when they cannot solve a question or problem on their own.

This collaboration between bots and agents is an effective way to provide complete and satisfactory customer service and optimize bot training.

How does ‘Ask for help’ work?

During the interaction with the user, the bot encounters a scenario where it needs to “raise its hand”. The (human) bot supervisor receives a notification indicating that it needs their assistance to resolve a case. This notification may be accompanied by relevant details about the user’s query and the context of the conversation up to that point.

Then, the supervisor provides the response to the bot so that it can continue interacting directly with the user and may choose to add that response to the knowledge base or not.

With this interaction, the bot can continue to learn and improve its ability to resolve similar queries in the future. This helps ensure that it handles an increasingly wider range of questions on its own over time, reducing reliance on human intervention.

Use case: a banking client owns a credit card which is no longer offered on the bank’s products portfolio

In the banking sector, it is common to launch limited edition credit cards, with benefits, agreements and exclusive rates that after a while are no longer offered but are still active for those customers who managed to acquire them during the promotion time.

In this specific use case, let’s imagine that a banking customer contacts the bot looking for information on discount agreements from sports stores with their Tokyo 2020 Olympic Games commemorative credit card. This product is no longer offered and was not included in the knowledge base when the bot was implemented.

The bot then requests help from a human agent via the Ask for Help feature. The agent provides the bot with the list of participating stores and brands but also takes the opportunity to remind the customer that the associated benefits will be valid until July 1, 2024, so it can offer alternatives available in the bank’s current portfolio that could be of their interest.

The supervisor decides to save the response and with this interaction, the bot learns about a specific situation that was not foreseen in its initial knowledge base. In addition, the customer receives the accurate and updated information they were looking for, which contributes to a satisfactory experience and could even result in a cross-selling strategy.

inConcert’s Ask for Help feature provides an effective solution for cases where bots face limitations in their ability to resolve queries autonomously. This collaboration between bots and agents not only optimizes the Contact Center’s operation but also enables the bot to learn and improve continuously.

With Ask for Help, a perfect balance is created between the efficiency of Artificial Intelligence and the human touch, promoting a stronger relationship between the customer and the company.

If you want some tips on how to implement this solution to your operation, do not hesitate to contact us.