AgenaRisk's Bayesian technology is based on 30 years of research in computer science, AI, Bayesian probability, statistics and smart data. It is designed for organizations that need to assess and manage risks in areas where there is little or no data, where direct measurement is not possible, and in the face of new, often novel, circumstances.
AgenaRisk's Bayesian Network technology is designed to model statistical data when relevant data is available; when it is not available AgenaRisk exploits domain knowledge. By fusing data and expertise, in the form a causal network model of the problem, it helps move from a 'big' data to a 'smart' data perspective. Computations are carried out on the Bayesian Network using state of the art algorithms. Unlike other technologies that require computer and data scientists to fit the model to data, AgenaRisk provides an easy to use, highly visual integrated design environment to build, test and deploy interactive risk models.
"Risk Assessment and Decision Analysis with Bayesian Networks"
The founders of AgenaRisk have produced a book based on 30 years of consulting experience.
We publish all of our methods and results and believe in open methodology. Our book is both a manifesto and a handbook of methods, models and pratical tips.
It has received excellent reviews on Amazon and is available now.
We believe that practitioners learn by example rather than theory. And whilst the theory is important (after all, the algorithms need to be correct) it is just as important to provide a helping hand.
All AgenaRisk products are delivered with an extensive library of models covering a huge number of application areas including: Project risk, Operational Risk, Stress Testing, Legal Reasoning, Medical diagnosis, Financial decision making, Value of information analysis and more!
DOWNLOAD FREE TRIAL
AgenaRisk Desktop is available as a free download. The full model library is available and access to functionality is unrestricted for the duration of the trial.
After the trial expires models from the model library can still be executed for research and educational purposes.