Skip to content

February 22nd, 2024

Webinar Replay: Level 4 Conversational AI: Redefining Customer Experiences

  • portrait of Kara Hartnett

    Kara Hartnett

Thank you to everyone who joined us for the insightful journey into "Level 4 Conversational AI: Redefining Customer Experiences." For those who couldn't attend or wish to dive deeper into the discussions, we've made the session available for on-demand viewing.

Featuring insights from Rasa experts:

Alan Nichol, Co-founder & CTO

Daksh Varshneya, Sr. Product Manager

Lauren Goerz, Sr. Product Marketing Manager

The focal point of our session was Rasa’s CALM technology, a revolutionary approach transforming conversational AI by enabling a more intuitive interaction between AI and users. This innovative technology marks the beginning of a new chapter in conversational AI, moving beyond traditional intent-based models to a more fluid, understanding-based model.

During the webinar, our experts guided the audience through:

  • An in-depth exploration of the five levels of conversational AI, with specific emphasis on how CALM propels organizations to achieve the fourth level.
  • A live demonstration of CALM, showcasing its capability to tackle conventional challenges associated with large language models in conversational AI.
  • An interactive Q&A segment where participants engaged directly with our experts, gaining insights and sharing perspectives on the future of conversational AI.
  • An introduction to the Rasa Pro Developer Edition, offering developers complimentary access to Rasa's cutting-edge innovations.

Stay ahead in the rapidly changing field of AI communication. Watch the replay today and be part of this transformative experience. Click here to watch.

Here are some questions we weren’t able to live answer in the webinar:

Q: How does the concept of Rasa “stories” tie into the “flow” concept?

A: While stories in Rasa helped developers by mapping out potential user interactions, they required anticipating every possible turn the conversation could take. Flows, introduced with CALM, simplify this process significantly. By focusing solely on the business logic required to complete tasks, flows remove the need to script every potential user input, streamlining development and making the bot's responses more dynamic and user-focused. This fundamental shift allows developers to concentrate on delivering value rather than micromanaging conversation paths.

Q: How does CALM adapt to low-resource languages especially for dialogue understanding in production use cases?

A: The flexibility of CALM's approach lies in its compatibility with various LLMs, which are pivotal for understanding dialogues. DIET, our robust NLU model, has empowered chatbots to converse in over 100 languages, demonstrating Rasa's commitment to linguistic diversity. With CALM, we aim to extend this inclusivity, leveraging the strengths of different LLMs to ensure that even languages with fewer resources are supported effectively in conversational AI applications.

Q: Are the scripted responses in general used word-by-word by the bot as defined in the flow? Are even the follow-up or clarifying questions by CALM scripted/ pre-defined?

A: At its core, CALM prioritizes delivering reliable and coherent interactions, initially relying on predefined responses to ensure accuracy and consistency. However, recognizing the need for conversations to flow naturally, CALM incorporates the Contextual Response Rephraser. This innovative component fine-tunes scripted responses based on the conversation's context, infusing each interaction with spontaneity and fluidity. It's this blend of predictability and flexibility that distinguishes CALM, making AI assistants more engaging and relatable to users.

Rasa Pro Developer Edition - Sign up here

For those ready to explore the transformative potential of CALM, the Rasa Pro Developer Edition offers a comprehensive platform to begin crafting your next-generation AI assistants. This is an opportunity to access the latest in AI technology and be at the forefront of conversational AI development.