2.3 Using the OFS Compliance Studio Application

This section describes the following key capabilities and concepts in detail.

Workspaces

Workspace Management is where the workspace Administrators define datasets and make them available to modelers in OFS Compliance Studio. It is an environment where the data is prepared for modelers to use in their modeling creation activity.

Workspaces are provisioned with the data required for modeling by administrators, who configure workspaces with subsets of production data. The data in the Workspace is made available to modelers using datasets, who then build models with the dataset without any exposure being provided to the physical data tables and columns in the database. In effect, the data is ready for the modelers, and they do not have to undergo the arduous task of accessing and querying the database. See the Workspaces section for information on how to use Workspace Management in the application.

Dataset

Datasets allow for the creation of a dataframe, a data structure that captures the logic that organizes data into a 2-dimensional table of rows and columns. Datasets allow for the reuse of these dataframes across models, as well as the features derived from the data source(s).

The Dataset allows you to manage the entire operation related to data set. You can perform the following two things using the Dataset window:
  • Define a metadata on how you want to create a dataset.
  • In addition, a mechanism where you can actually take a snapshot of real time data and store it. So, it can be used later in the pipeline.

    For more information, See the Dataset section for more information.

Pipelines

Modeling refers to the process of designing a prototype based on a structured data model for statistical analysis and for simulating actual events and functions. A user with access to the Workspace can create or modify models in a workspace. Model versions are preserved in the Workspace, along with execution and output histories. Once a model has been validated in the Workspace and considered fit for use, modelers can request to push the Model into the production environment.

See the Pipelines to create models from pre-defined models to predict business trends and validate the existing models with the help of Compliance Studio.

Model Actions

The Model Actions window allows you to view the list of models associated with the Workspace.

For more information, see the Model Actions section.

Graphs (Graph Pipeline)

Unlike traditional relational database management systems, the Graph Pipeline feature allows you to view the data relationships in a graphical format.

This feature uses the latest technology to harness the power of Graph Analytics to give Financial Institutions the ability to monitor the data financial institutions effectively. The data is organized as nodes, relationships, and properties (property data is stored on the nodes or relationships). The results of analytics algorithms are stored as transient properties of nodes and edges in the Graph.

For more information, see the Graphs section.

Scheduler Service

The Scheduler Service is a service in the Infrastructure system that automates behind-the-scenes work that is necessary to sustain various enterprise applications and functionalities. This automation helps the applications to control unattended background jobs program execution.

The Scheduler Service contains a graphical user interface and a single point of control for the definition and monitoring of background executions.

For more information, see the Scheduler Service section.

Audit Trail

The Audit Trail window provides the complete details of model. This shows the information such as, when Model was created, who created the Model, workflow of Model, for example when this Model became champion or deployed, and so on.

For more information, see the Audit Trial section.

Data Studio Options

Compliance Studio contains an underlying Notebook Server which has the following configurable options:
  • Interpreters
  • Tasks
  • Permissions
  • Credentials
  • Templates

For more information, see the Data Studio Options section.

Object Migration

Object Migration is the process of migrating or moving objects between environments.

You may want to migrate objects for several reasons such as managing global deployments on multiple environments or creating multiple environments so that you can separate the development, testing, and production processes.

For more information, see the Object Migration section.

Model Libraries

The Model Library information is used to bind a particular technique and its details to one unique Model Library. The Starting point for the Model builder is to register a model library with application after it is properly set up in the python environment to be used.

For more information, see the Model Libraries section.

Model Techniques

Model Technique is the algorithm/technique used to create python model using the library/package which was created in the Model Library Screen. It is the actual information captured in the application that helps in training the model (Upload and Build).

For more information, see the Model Techniques section.

Model Catalog

The Model Catalog page allows you to add and manage the Model Technique, Model Library, and Model Objectives. You can either import the models from external sources or create, train and label it as champion model.

For more information, see the Model Catalog section.

Ruleset Details

The Ruleset facilitates identifying the similarity between two entities (customer, account, etc) and derives a match. A Ruleset is a set of rules applied to the defined source and target entities and compares the entities' attributes to derive a match.

For more information, see the Creating Rulesets section.

Manual Decisioning

The Manual Decisioning in Compliance Studio allows you to make manual decisions on similarity edges from the threshold set on the Match Ruleset page. This flexibility provides an Ad-hoc experience to improve the graph analytic capability and identify the match to analyze risk and take required actions.

For more information, see the Using Manual Decisioning section.

Merge and Split Global Entities

Merge and Split Global Entities in Compliance Studio allows you to manually break, merge, create, and re-arrange parties into global parties. This flexibility provides an Ad-hoc experience to enable users to make decisions on borderline cases or where data issues or rules logic has grouped records incorrectly.

For more information, see the Using Merge and Split Global Entities section.

ML4AML Model Templates

The prepackaged Model template allows you to perform the following:
  • Customer Risk Scoring
    • Model segmentation (model grouping)
    • Load and Preview data
    • User-defined transformation (deriving additional features)
    • EDA
    • Feature selections
    • Model training and evaluation
    • Model Agnostics (Explainability)
    • On-going validations
  • Customer Segmentation
    • Anomaly detection (within Cluster)
    • Deviation detection (across Cluster)
  • AML Event Scoring
    • Train and Score ECM events
    • Classify events into Low, Medium and High
  • Typology Scenarios
    • Shell Account detection
  • Automated Scenario Calibration
    • Prod scenario analysis
      • Scenario Execution
      • Above the Line (ATL) Analysis
      • Below the Line (BTL) Analysis
      • Outcome/Impact Analysis
      • Transaction Analysis
    • Pre-prod Scenario Analysis
      • Scenario Execution
      • Pre prod Analysis
      • Transaction Analysis

For more information, see the OFS Compliance Studio ML4AML Use Case Guide.