Hello, i’m working on a project based real time data monitoring using LLM, the idea is to create a tool ( can be a gpt or an app ) to answer questions based on daily dashboards on kibana elasticsearch, the agent or the assistant (or whatever can be called…) should be able to extract anomalies from sales dashboards from different countries, each dashboard contains 2 sales signals, one for the day of obseravtion and one for average sales of 7 days before ; if the sales increase/decrease from day of observation by 50 % compared to the signal of 7 days before , the agent should be able to detect it, can assistant be a solution for this ? or i can use other alternatives, i will appreciate your guidance.
Thank you
You could create an API to identify anomalies in the dashboards and generate relevant content that can be sent to LLM to produce the final response. It’s not too complicated.