Create and manage multiple OpenAI Agents with image upload and database storage Beta Version of streaming assistants api creator and manager. The app runs in streamlit and allows the user to create and manage multiple assistants in a single session. The build defaults to the latest non reasoning openAI LLM. The app allows the user to upload images and have the assistant use the image data to perform tasks designated as instructions. The threads can be maintained between chat sessions. The database is a very straitforward SQL lite that runs in the Google Cloud. The entire coding effort is meant to be a testbed for learning to interface with the OpenAI API platform. The app serves as a virtual playground where multiuse assistants function together on a single thread. The entire premise is to test the state of the art related to leveraging the inherent emergent abilities and behaviors of the latest LLM and meshing that with a coding infrastructure that provides very practical tools for the user. The idea is also to test the possibility of seeding inherent agentic processes by multiple assistants acting in concert on a user input data or images. Testing more extensive and iterative large instruction sets allows the user to create novel AI assistants that may nurture agentic collaboration.

working website for the app => https://medbotsolutions.com

email me at sbayer2@gmail.com if you have questions or improvements