Epistemix’s Framework for Reconstructing Epidemiological Dynamics (FRED) is an agent-based modeling platform for simulating how conditions spread through populations. Modelers use FRED to figure out how and why particular conditions—diseases, ideas, behaviors, attitudes, economic status, etc.—impact communities across space and time.

There are many different approaches to building computational models, and we built FRED to inform specific policy interventions. Unlike other modeling strategies that are limited to projecting high-level trends, FRED is most useful when used to simulate and compare the consequences of specific key decisions like a state governor adjusting COVID-19 social distancing policies, a hospital forecasting ICU needs, or an enterprise manager balancing operational risk.

epistemix visualization dashboard

Realistic Social Dynamics

FRED captures the social, environmental, and behavioral elements of individuals, making it a unique tool to model complex social contagions and epidemics linked to social determinants.

FRED can simulate specific interactions among millions of individuals in a large population. Each individual person is computationally represented, simulated, and tracked, and behaves differently depending on its particular predispositions and local circumstances.

Prebuilt Synthetic Population

In order to provide decision-makers with simulations of epidemics in particular locations, FRED features an extensive, modular, high-quality synthetic population for the United States that represents individuals, their statistically accurate households, neighborhoods, schools, and workplaces, and demographic characteristics including age, sex, and race.

Population and Individual Level Output

FRED keeps a condition record for each individual agent over time so that modelers can access granular data about any given scenario.

Computational Efficiency

FRED limits as much computation as possible to the active set of individuals and those individuals who may directly interact with the active set. Other agents are updated on a strictly as-needed basis. For models in which a small subset of a large population is in the active set, FRED can produce dramatic speedups.


FRED models can track any number of interacting conditions spreading through a population.

Bottom-Up Modeling

FRED enables you to study how individual behavior gives rise to macro-level phenomena by accounting for heterogeneity of populations, individuals' autonomy and bounded rationality, interactions that lead to emergence across populations, and the explicit space where interactions happen and space agents occupy.

ArcGIS Integration

Import FRED data into the suite of Esri ArcGIS products (ArcGIS Pro, ArcGIS Online) to create maps, solve problems, and share geospatial data to preform location-based analyses and gain greater awareness.


Because FRED’s synthetic population represents individuals that are statistically correct, rather than actual people, it can model populations without infringing privacy. Supplemental client data sets can be processed to remove any identifying information.

Learn how health care systems, governments, enterprises, and researchers are using the FRED platform to respond to COVID-19.