This document introduces Advanced Temporal Language Aided Search (ATLAS), a new interface for defining clinical phenotypes using natural language. ATLAS allows users to search for and write phenotype definitions step-by-step rather than using drop-down menus. The document demonstrates how common phenotypes like type 2 diabetes and acute myocardial infarction can be defined using ATLAS. It reports that a group was able to define 38 phenotypes in around 2 hours using ATLAS, much faster than previous rule-based methods. The Common Data Model may be extended in the future to additional support for ATLAS, such as time attributes and clinical note annotations.