I have a requirement to allocate a server(VM) for each user request and make this automatic.

Is it possible with Openstack autoscale? 

Scaling should not be done based on either cpu or memory usage as only one user is supposed to access a single VM.

Regards,

Danishka