this should be kept in the async decision layer only
1.dynamic parameter tuning adjust grids,eta weights ,etc
2.demand supply intelligence agent - basically identifies hotspots ,dip zones ,etc
3.can create an agent that can be used for predefined strategies like supplying drivers in zones based on expectation of spikes in customer count
the road map will be like
kafka ----> ai agent ---> redis
this makes sure that no effect on latency by the agent
this should be kept in the async decision layer only
1.dynamic parameter tuning adjust grids,eta weights ,etc
2.demand supply intelligence agent - basically identifies hotspots ,dip zones ,etc
3.can create an agent that can be used for predefined strategies like supplying drivers in zones based on expectation of spikes in customer count
the road map will be like
kafka ----> ai agent ---> redis
this makes sure that no effect on latency by the agent