I’m developing a Gradio-based interface where each user can submit a request that initiates a CrewAI flow. Previously, for single-user scenarios, I wrote the user request to a JSON file and triggered the CrewAI flow using a subprocess, which read the JSON for input parameters. Now, as I scale to support multiple users, I need to pass additional information - the username, along with the request, so that the process will happen independently for each user. What is the recommended approach to manage and pass these user-specific inputs to CrewAI flows in a scalable manner?
I’m assuming you have authentication on the gradio application, you should be able to get the user details in the request object in the function you use to kickoff the crew.