When we started work on the idea for Alana, we considered everything human conversation gives us. It answers questions. It provides companionship. It builds connections, shows understanding and allows us to explore and discover new information. While existing task-oriented AI assistants can participate in a question-answer dynamic, they can’t perform any of these other vital functions.
We created Alana to start a more multi-dimensional AI conversation.
When conversations stray off script, Alana listens and follows. It has the capability to decipher misunderstanding, engage with small talk and gloss over awkwardly articulated commands and questions. So, in whatever direction a conversation meanders, Alana understands, responds and keeps an extended dialogue flowing forward.
By learning from real conversations, we use machine learning to embed natural, human-like behaviour into AI conversational tools. Alana recognises the subtle conversational cues that signal where a line of chat is heading. A pause that suggests indecision.
A change in tone that indicates satisfaction. Or an expression that hints at an upturn in mood. It makes connections, and takes what it learns from interactions to proactively signpost users to information that matches their interests.
By remembering previous conversations, Alana can shape future dialogue and personalise exchanges. Memory is a powerful addition to AI conversational technology. It allows Alana to make associations, identify correlations and then use this knowledge proactively to make suggestions. This makes Alana much more than your standard AI assistant. It creates a companion.