Agena Company Logo

Agena Company Logo
Sales Enquiries

If you are interested in buying AgenaRisk please get in contact with us here.

AgenaRisk Free Download
You can download a free version of AgenaRisk   here.

AgenaRisk Lite Download

If you have purchased a copy of Fenton and Neil's book you can download your copy of AgenaRisk Lite  here.


Fenton and Neil's book: "Risk Assessment with Bayesian Networks"

BN Book

AgenaRisk users include:

Customer Logos

You are here: Home » Products » AgenaRisk Desktop

AgenaRisk Desktop


AgenaRisk is a powerful but intuitive tool for modelling, analysing and predicting risk. If you have to make decisions involving uncertainty as part of your work, you will find AgenaRisk to be an invaluable aid. AgenaRisk supports both diagnostic and predictive reasoning about uncertainty using risk maps, otherwise known as Bayesian networks.

AgenaRisk has the following key features. Click each screenshot to see a full-size version:

Risk Maps (Bayesian networks) for modelling causal relationships.
Probability Tables that allow you to quantify these relationships using either expressions or explicit probability values.
Risk Tables for entering data quickly and efficiently as you would using a spreadsheet.
Highly configurable Risk Graphs that display the results of running your models.
A Risk Explorer that allows you to build and view complex, nested models.
Simulation of a wide range of arithmetic functions and statistical distributions.
A Sensitivity Analyser that allows you see the sensitivity of nodes based upon a variation of sources. mv analysis
A Mutivariate Analyser that allows you to export bivariate distributions, involving pairs of nodes, along with their accompanying correlation (Pearson and Spearman) statistics and summary statistics.  mv analysis
A large repository of Sample Models that cover different modelling problems.


The state-of-the-art algorithms implemented in AgenaRisk allow you to model the following classes of problem:

  • Simulation of statistical distributions for predictive inference
  • Diagnostic inference for machine learning applications
  • Hierarchical modelling as an alternative to Monte Carlo Markov Chains (MCMC)
  • Construction of hybrid models containing discrete and continuous uncertain variables
  • Mixture modelling of discrete and continuous distributions
  • Representation of expert judgement using subjective probability
  • Dynamic modelling of time-based or evolving systems (e.g. Markov analysis)
  • Object-oriented modelling of complex systems involving multiple objects and interfaces

AgenaRisk has the following features:

  • Risk Maps (Bayesian networks) for modelling causal relationships
    • Point-and-click functionality for creating models quickly and easily
    • Fully customisable appearance of nodes and edges (shape, colour, border, text)
    • Grouping and alignment
    • Copy and paste
    • Zoom
    • JPEG import and export facility
    • Support for adding customisable labels and boxes
    • Optional display of risk graphs on nodes
    • Ability to hide nodes and edges
    • Intuitive interface for Object-Oriented Bayesian Networks
    • Comprehensive node typing system
    • Pre-defined ranked nodes to simplify creation of certain types of complex risk model
    • Numeric nodes for unprecedented accuracy
    • Powerful editor and wizard for adding states quickly and easily
    • Model import

  • Probability Tables that allow you to quantify these relationships using either expressions or explicit probability values
    • Multi-dimensional tables for editing probabilities directly
    • Copy and paste
    • Automatic normalisation of probabilities
    • Powerful expression editor with wide range of pre-defined statistical and deterministic functions
    • Support for expressions partitioned over parent state combinations

  • Highly configurable Risk Graphs that display the results of running your models
    • Support for bar graphs, line graphs, scatter plots and histograms
    • Multiple data sets shown simultaneously to make comparison easy
    • Fully customisable appearance
    • Optional display of statistics on graphs (mean, median and percentiles)
    • Summary panel showing statistics and actual state probabilities
    • Support for cumulative plots
    • Intelligent truncation and scaling to focus on interesting areas of data
    • Different configurations for small and large data sets
    • Support for discrete or continuous plots
    • Flexible display of graphs (in a dedicated panel, in separate windows or on risk maps)

  • Risk Tables for entering data quickly and efficiently as you would using a spreadsheet
    • Fully configurable names, answers and descriptions of entries
    • Support for creating new headings and reorganising tables
    • Easy creation of user-friendly risk assessment applications
    • Synchronisation of entries with nodes in risk map
    • Colour-coded cells to show data values at a glance
    • Display of multiple scenarios side-by-side for easy comparison
    • Import and export of scenario data

  • A Scenario manager which allows you to create multiple scenarios, clone, delete and show/hide visibility in Risk Graph/Table. Multiple scenarios give you the capability to compare data in both Risk Graph and Table views.
  • A Sensitivity analyser which allows you to analyse sensitivity of nodes based upon a variation of sources
    • Select multiple node sources to visual sensitivity of a target node
    • For continuous nodes, select summary statistics as choice of data to analyse sensitivity on
    • Select any or all of calculated data table, ROC curves and tornado graphs as your choices of output
    • Preview sensitivity analysis report in your web browser
    • Export sensitivity analysis report to html
  • Table learning from data and expert judgment  This enables you to automatically learn node probability tables from a spreadsheet file and incorporate as much or as little expert judgment as you wish.  In all cases learning can be performed even when there are ‘missing data’ in the spreadsheet. The learning process is performed using an Expectation-Maximisation (EM) algorithm 

  • Import/Export functionality that allows you to import/export data from your models

    • Export your data to CSV files and import after editing
    • Open the data file in a spreadsheet or other CSV supporting application for editing
    • Export your models as a whole or individual risk objects

  • A Risk Explorer that allows you to build and view complex, nested models
    • Intuitive view of nested models that can be expanded and collapsed
    • Support for importing new models
    • Support for Object-Oriented Bayesian Networks

  • Accurate Simulation of a wide range of arithmetic functions and statistical distributions

    • Dynamic calculation of appropriate intervals
    • Fully configurable settings for achieving the right balance between performance and accuracy
    • Arithmetic functions:

      • Addition
      • Subtraction
      • Multiplication
      • Division
      • Power
      • Modulus
      • Logarithm
      • Natural Logarithm
      • Square Root
      • Modulus
      • Min
      • Max
      • Weighted Mean
      • Weighted Min
      • Weighted Max
      • Min-Max Mixture
      • Equals
      • Not Equals
      • Less Than
      • Less Than Or Equals
      • Greater Than
      • Greater Than Or Equals
      • NOT
      • AND
      • OR
      • XOR
      • NoisyAND
      • NoisyOR
    • Statistical distributions:

      • Beta
      • Binomial
      • Chisquare
      • Continuous Uniform
      • Exponential
      • Extreme Value
      • Gamma
      • Geometric
      • Hypergeometric
      • Integer Uniform
      • Logistic
      • Lognormal
      • Negative Binomial
      • Normal
      • Poisson
      • Student-t
      • Triangular
      • Truncated Normal
      • Weibull

  • Compound sum analysis This enables you to automatically compute the compund  'total value' distribution where there are events with a frequency distribution, and where each of these events has an associated severity distribution. 
  • A large variety of Sample Models that cover different modelling problems

    • Aggregating Distributions
    • Asia
    • Batch Learning Model
    • Biased Coin Flip Experiment
    • Car Costs
    • Causal Risk Register
    • Constraint Satisfaction
    • Dependent Coin Flips
    • Diabetes Treatment
    • Diet Experiment
    • Fault Tree Analysis
    • Fire
    • Headache
    • Hypothesis Testing
    • Intensive Care Monitoring
    • KUUUB
    • Mildew
    • Mixing Product Failure Data
    • Monty Hall Dilemma
    • Mountain Pass
    • Naive Bayesian Classification
    • Operational Risk in Finance
    • Printer Fault Diagnosis
    • Reliability Estimation
    • Risk Control Self Assessment
    • Risk Drivers and Indicators
    • Safety
    • Simple Testing Process
    • Six Sigma Defect Containment
    • Six Sigma Testing and Rework
    • Statistical Distributions
    • Water Purification
    • Wet Grass

System Requirements


  • 1 GB RAM minimum (actual memory demand depends on size of model)
  • 30-80 MB hard disk space (depending on o/s and bundled installation of Java JRE)
  • 2.4 GHz processor minimum

Operating System details:

  • Microsoft Windows  10, 8, 7, Vista, XP (with Java JRE 1.8 bundled)
  • Macintosh 10 (uses default Java JRE installed on machine)
  • Ubuntu (Linux/Unix) (uses default Java JRE installed)
  • AgenaRisk runs best with JRE 1.8.
  • There are 32-bit and 64-bit versions of AgenaRisk. 

How To Buy

For pricing contact

For general technical questions about AgenaRisk as well as many hints, tips and videos, join the AgenaRisk LinkedIn User Group.
Maintenance Policy

AgenaRisk is now sold on a subscription basis only. For the duration of your subscription, you are entitled to:

  • Free technical support via email. Technical support is available for software installation, resolving software errors and assisting with software operation. Technical support is not designed for building models from scratch, model de-bugging or AgenaRisk training. These services may be obtained separately from Agena’s Training and Consulting team.
  • An unlimited number of incidents.
  • Guaranteed response within one business day.
  • Free maintenance updates and major upgrades of AgenaRisk.