A conversation search on fashion products and personalized results. Example:
How it works? -Inititally insert the image/text clip embeddings of products in the Pinecone database.
-Gathering insights from prompts,user history using key value pairs generated by gpt-4o-mini.
-Use the key value pairs to filter and search most similar product in the pinecone database.
##Tech Stack
-Langchain Openai for prompt,history processing.
-Clip model for embeddings,Pinecone for vector database.
-FastAPI for backend.
-React for frontend.