Posted On: Nov 19, 2020
We are excited to announce the availability of Context Management on Amazon Lex. Conversations involve managing context across multiple turns as the interaction evolves. Similarly, bots need to understand the progression of a conversation’s context in order to respond appropriately. Previously, you had to write code to manage context via session attributes. Depending on the intent fulfilled, the code had to orchestrate invocation of the next intent. Starting today, Lex supports context management natively so you can manage the context directly from the console. With the context management capability, you can easily control when an intent should be activated. For example, consider when a user asks “What were my expenses this month?” and then maintains the same context relating to expenses in the subsequent turn “How about last month?” Utilizing the context natively allows you to create sophisticated, multi-turn conversational experience without having to write any code. In addition, it is now possible to set default slot values. You can set the default to a constant, an active context attribute or a session attribute.
Context management along with default values augment the Amazon Lex design toolset to help simplify the conversational design of your bot. Now you can quickly add context into the conversation flow and increase the effectiveness of your virtual agent, conversational IVR system, self-service chatbots, or informational bot.