Prompting In MEVA
Prompts are predefined messages or questions used in MEVA to guide user interactions and prompting the user to provide specific details or responses needed to continue the dialogue or achieve a particular goal. They help structure the conversation, ensuring that MEVA collects necessary information and provides relevant assistance. Prompts can be tailored to various scenarios, such as asking for user details, offering choices, or providing instructions. Prompting can be done in 2 ways, one is a pre prompt and the other is a post prompt.
Pre-Prompt
A "pre-prompt" is like a guiding statement or instruction given to the MEVA before it starts generating text. This initial input helps set the context and direction for the MEVA\'s response, making it more relevant and accurate. By providing this guidance, the pre-prompt ensures that the MEVA understands what you want it to focus on, leading to better and more useful responses.
For example: For a welcome intent,
"You are MEVA AI for a support platform, your goal is to greet users and welcome them warmly. Ensure your tone is polite throughout the conversation."
With this instruction, MEVA understands that it has to welcome the user with a warm and polite greeting.
Post-Prompt
Post-prompt refers to the actions and responses that occur after the user inputs a prompt. Once the prompt is received, the MEVA analyzes the text, understands the context, and generates a relevant and coherent response based on its training data. This process involves interpreting the input, considering various possibilities, and selecting the most appropriate reply.
For example: In the welcome intent, instruct MEVA on what it can do after the user has messaged,
"After delivering the welcome message, ensure to provide options or guide the user on how they can proceed. Offer examples of common inquiries or actions they can take next. Keep the tone friendly and encouraging"
Post-prompt processing ensures that the MEVA provides accurate, helpful, and contextually appropriate answers, enhancing the overall user experience and effectiveness of the conversational AI. This stage is crucial for maintaining the flow of interaction and ensuring the user\'s needs are met effectively.