ICONTEK replaces ineffective single swoop automation with a hybrid – human plus machine – operating model. The transition to fully automatic machine-led operations occurs in three low-risk stages, allowing operators to train the engine in real-world environments without risking any poor user experiences.
End users have their calls answered by a bot that has a library of responses to give based on the company’s dialog decision tree (DDT). Since agents don’t need to speak, they transition from speaking agents to silent agents. In stage I and II – training and verification – human agents and machine-learning bots operate in unison, while human agents have final control over actions taken. Then, in stage III, the bot operates independently.
Bot learns by observing silent agents and end users. The bot does not make any decisions. Human agents control the bot’s actions by selecting responses from a menu.
Bot actions are verified by silent agents. The bot is making its own decisions, while a human agent reviews and approves (or changes) decisions before action is taken.
Bot makes decisions and takes actions . The bot is let loose in the wild and runs independently, making decisions and taking actions without human agent assistance.
End users have their calls answered by a bot that has a library of responses to give based on the company’s dialog decision tree (DDT). Since agents don’t need to speak, they transition from speaking agents to silent agents.
In stage I and II – training and verification – human agents & machine-learning bots operate in unison, while human agents have final control over actions taken. Then, in stage III, the bot operates independently.