How would you refine risk models, explore alternative datasets, or optimize execution algorithms at Aqr?
