The State Changers deliberated on the challenges over the integration of Dialogflow with Zano. Due to unsuccessful results, a detailed analysis of the API and UI was carried out. The main focus of the meeting was on the enhancement of the process, specifically on how the conversation flow can be advanced based on the information provided by users.
The main issues identified were: 1. The AI's inability to understand clear answers to Yes/No questions, treating 'No' as indeterminate intent, without any fallback mechanism. 2. Absence of suitable handling of negative user intent. 3. Manual handing of API responses is inefficient and guides users to an unintended page. The State Changers concluded that rather than attempting to overhaul the system totally in Zano, it may be more beneficial to address the identified challenges at different points individually. For instance: 1. Inclusion of a 'No' inference in Dialogflow to handle negative user intent. 2. Utilization of event handlers in Dialogflow to manage 'No input' or 'No match' scenarios by providing specific responses back to the user. 3. Parsing of the requests and responses more effectively to understand the context of the conversation and adjust the flow accordingly. These small modifications at different points may potentially lead to a vast improvement in the system performance overall. The need for further tests and necessary adjustments was recognized.
(Source: Office Hours 7/5 - Handling Unhappy Paths, Adding more custom views in Bubble )
