Short bio

I am a modeler with expertise in economics, politics, security, statistics, operations research and data mining. I specialize in building analytical tools, decision support systems and multi-agent simulations. I have a Ph.D. in Computational Social Science from George Mason University and a M.Sc. in Quantitative Methods from the Warsaw School of Economics

I co-founded Scensei, a decision support and analytics enterprise. I was a Research Assistant Professor at the Krasnow Institute for Advanced Study. Currently, I am CTO for positif.ly and trovero.io

Mission statement

Multi-agent simulation is the core of the analytical and decision support tools I build to help decision makers to gain insight into the problems they face; to design policies, plans, procedures and programs  and to evaluate their effectiveness.

A model can be expressed as a narrative, for example some text that details why a national leader decides to go to war with a neighboring country; or as equation, for example, a regression or system dynamics model that shows the link between incidents of suicide terrorism in a country and levels of poverty in it. It can also be expressed as computer code. One way of coding a model is to create artificial humans, their purposeful behavior and the environment they live in. These virtual humans are called agents. Since a model often has many interacting agents, it is called multi-agent. Agents interact with each other as their human counterparts do, but in a somewhat simplified manner. To have agents interact, we need to simulate the virtual world we create. That is why multi-agent models are also called multiagent simulations.

High-fidelity multi-agent simulations use valid and detailed information on the environment and the behavior of the humans they represent. Resolution of the model in terms of it's data fidelity, scale, behavioral and cognitive plausibility needs to be matched to the question on hand. This is my craft.