Task

  • Chat modality, where the user engages in a conversation with the AI so the AI can create the most hyperpersonalized and richly textured painting - this conversation happens through a chat window
  • The user chats (through voice modality or chat modality) with the AI about a certain theme, idea, project, set of emotions, problems, issues they have been grapping with.
  • Say they have carried a deep regret in heart, or some ligering pain from a heartbreak, or there is a challenge ahead of them for which they seek deep reserves of resolve and inspiration, or they want a “positive” emotions focused painting, with themes like hope, conviction, inspiration, beauty, joy, wonder, due to something they expect to happen soon - say the arrival of a child, or an upcoming adventure or trip of a lifetime.
  • The AI and the user talk about this theme or topic, for a minute or 10. Once the conversation is completed, this is transcribed (if in text modality) fed into an LLM to be converted into a detailed, comprehensive, richly textured and nuanced, hyperpersonalized, prompt, where none of the key details are missed from the conversation. This prompt is then fed into an image model/multimodal model. The results (3 options) are returned to the user.

plan of action

conversation that goes deeper about a topic

  • text and voice both required (use livekit maybe)
  • system prompt

llm call that takes a conversation and outputs image


Links : TODO

Tags :

Date : 11th April, Friday, 2025, (Wikilinks: 11th April, April 25, April, 2025. Friday)

Category : Others