Gen AI with Journey orchestrated actions

What it does

This is a banking example using streamlit, Gemini and Journey to orchestrate sending of actions. The chat application uses gemini to classify what user is asking, determines privacy level of the request, and determine which pipeline action to perform, then uses gemini to generate the json call to trigger the action to authenticate, or trigger the action. In this video example, we answer a couple general banking questions, then get into asking for our account balance where the gemini recognizes the privacy level and triggers out the authentication action pipeline (from Journey) and once authorized, later asks to make a payment which is then also triggered and handled by journey. This is a zero knowledge network orchestration of AI with general information requests and actions. The AI is generating the action JSON as a function, as well. This is an MVP for this demonstration purpose, but ultimately, can be made fully dynamically driven with intelligent prompting in addition to GenAI capabilities of Gemini.