Our Technology

We combine proprietary IP for rapid cohort creation with automated, state of the art data science

Platform Overview

The Atropos Evidence Platform turns health data into publication grade evidence in less than 24 hrs with our unique combination of technology and highly trained staff.

How It Works

Step 1: We convert medical data stored in common relational data models into our proprietary in-memory database that stores data as patient timeline objects for faster cohort building.

Relational Data Model (e.g., Epic Clarity or OMOP CDM)

Proprietary In-Memory Data Model

Step 2: Cohorts of patients are selected using our domain-specific temporal query language, TQL.  These queries perform 50x faster and 30X cheaper than standard cloud relational databases (e.g., Google BigQuery).

Atropos Cohort Engine (ACE) Low-Code User Interface

Step 3: Query outputs feed our automated analytics pipeline to build user-ready reports and key artifacts in seconds.

(Representative artifact examples below)

Kaplan-Meier Survival Curve


Cohort Demographics Summary


High-Dimensional Propensity Score Matching Table

Propensity Score.png