ArrowMiner. User Roles

User Roles are permission sets that control access to areas and features within an ArrowMiner application. User accounts may have different roles in different applications assigned by either the application Administrator(s) or the application Owner.

A user account may have only one role per application.

Standard User Roles
  • User. A role granting access to the basic application features but without administrative access.
    There may be multiple Users assigned to an application.
    • Can run all application modules without restriction
    • Can create new decision records and apply decisions
    • Can double-check and approve decision records (4Eyes configuration)
    • Can comment any decision record
    • Can document open decision records
  • Supervisor. A role granting access to all application features with specific supervisor rights
    • Access to all User role features
    • Can apply decisions and override restrictions
    • Has access to specific supervisor features
    • Does not have access to administrative features
  • Administrator. A role granting access to all application features with administrative rights
    • Access to all User and Supervisor role features
    • Access to application settings
    • Does not have access to upgrade/downgrade or billing options
    • Cannot assign/modify an Owner role
    • Cannot assign/modify a role on himself.
  • Owner. A role granting access to all application features with administrative rights
    • Access to all User, Supervisor and Administrator role features
    • Access to all upgrade/downgrade or billing options.
    • Cannot assign/modify an Owner role
    • Cannot assign/modify a role on himself.
User Teams

Users may be organized in teams. Teams of users may have specific rights or restrictions on specific application features that override role rights, like:

  • Create new decisions
  • Apply decisions to existing open records
  • Perform Batch Uploads
  • Apply Decision Outcomes
  • Delete Records
  • Exclude Records..
Get Information On User Role & Rights
1 Select an application
2 Select the Application Info option in the ArrowMiner menu.
3 Get information on the active user rights
  • Application Owner. The application owner user name
  • 4Eyes Control. The application requires 4Eyes control on all decisions.
  • User Id. The active user id
  • User Role. The active user role
  • Decision Team. The active user’s team
  • User Can Create Decision?
  • User Can Apply Decision?
Get Information On application Users
1 Select an application
2 Select the Application  Users in the ArrowMiner menu.
3 The system willl display a table with the application users, the role of each user and a mailto link on each a user id.

Contoso-KYC (Demo Application)

The sample application describes a small-size private bank with approx. 4bios CHF under management, 2000 customers and a Compliance department composed of 4 officers.

The on-boarding process for new customers includes a pre-check phase (go, no-go) where a decision is taken on whether the process should go on or the customer should be rejected. The Compliance officer performs some preliminary checks on customer’s attributes like: source of funds, source of wealth, beneficial ownership, politically exposed persons, purpose of the account, services or corporation location that are added to some fundamental data like: assets under management, nationality, previous banking relationship in order to provide a first appreciation on the customer’s risk before a further investigation is performed.

Contoso suffers from labour intensive operations due to extensive manual processes, low quality case outcomes due to fragmented data and insufficient harnessing of technology, data and analytics. These shortcomings have damaging repercussions. They can result in low quality decision-making, costly remediation and regulatory fines, impaired customer experience and reputational damage.

Contoso’s goal is to shorten the overall customer admission process and higher the quality of the decision-making.

Contoso will use the ArrowMiner© and Know Your Customer application to significantly reduce the pre-check time, reveal very high risk customers, reduce associated costs and concentrate on the most relevant candidates.

The risk-based approach of ArrowMiner©, will allow Compliance to better distribute the work among different seniority levels within the team and produce a qualitative risk-based report.

ArrowMiner© is a Decision Support system associated to a risk-based methodology will use the organization’s risk appreciation and investigator’s decisioning process to learn from training data, classify the information and finally predict decision on future cases.

The Risk-Based classification functionality of ArrowMiner©, preceding the prediction, requires that the elements affecting the investigator’s decision process are thoroughly identified and measured:

  • Decision Attributes
    Decision properties that affect the human decision
  • Attributes Risk Rating
    Risk Rating scale of each decision attribute
  • Attributes Risk Weight
    The impact of each decision attribute on the human decision
  • Decision Rating
    The rating for each decision taken
System Configuration

Decision Attribute

 

Risk Rating

 

Risk Weight

Banking Relationship
The relationship type the customer has with the bank and the type of services that will be u̇sed. E.g. Retail banking customers are considered less risky than Private banking customers.

 

0-3
In-house rating : low to very high risk

 

5

Citizenship & Immigration
Customer’s country of origin or passport country.  Local customdesign are considered as less risky.

 

0-4
In-house rating : low to very high risk

 

5

Activity
Customer/Entity business activity

 

0-4
In-house rating : low to very high risk

 

3

Beneficial Ownership
Degree of transparency of ownership. Complex ownership structured are considered as more risky.

 

0-4
In-house rating : low to very high risk

 

3

PEP
Politically exposed persons should be handled with extreme care.

 

0-2
0=No PEP. 1=Indirectly PEP. 2=Directly PEP

 

3

AML Schema
Suspicious activity indicative of potential money laundering behavior

 

0-4
In-house rating : low to very high risk

 

5

Business Organization Type
The risk rating of the business activity. E.g. Casinos or Money Exchangers are considered as high risk activities.

 

0-3
In-house rating : low to very high risk

 

3

Source Of Wealth
This is a rating internal to the bank.

 

0-3
In-house rating : low to very high risk

 

3

Legal Form Structure
Legal form structure and risk associated. E.g. Sole proprietorship are considered as more risky.

 

0-4
In-house rating : low to very high risk

 

2

Financial Statements
The order rating of financial statements especially for credits or extensions of credits

 

0-4
In-house rating : low to very high risk

 

2

Source of Funds
Transparency on funds source. Where the initial funds come from.

 

0-5
In-house rating : low to very high risk

 

1

Transaction Risk Rating
The risk rating on intended transactions

 

0-3
In-house rating : low to very risk

 

2

Decision Rating
Based on internal customer risk rating and decision making.

 

0  Very Low Risk
1  Low Risk
2  Medium Risk
3  High Risk
4  Rejected

 

Note that all case attributes and decisions are rated following a strict risk-based approach reflecting the organization risk culture and eliminating any security issues relating to data exchange with the ArrowMiner© server.


Decide & Predict

This is the form for the new decisions on new customers. Note that the screen is composed by the: Ribbon (horizontal) above the form and the form tabs.

Ribbon

  • Expand/Collapse
    • This is a toggle button that will allow you to focus on the current tab.
  • Decision Prediction Based On Based On Previous Decisions On Similar Risk
    • Runs a decision prediction for the current case based on previous decisions and similar risk. This function uses the corporate memory and the decisions taken by the past. The function provides only indicative results.
  • Decision Prediction Based On Previous Outcome On Similar Risk
    • Runs a decision prediction for the current case based on previous outcome (not decisions) and similar risk. This function uses the corporate memory and the decision outcome in past decisions. This function will require an estimated outcome. The function provides only indicative results.
  • Outcome Prediction Based On Previous Outcome On Similar Risk
    • Runs a decision outcome prediction for the current case based on previous outcome (not decisions) and similar risk. This function uses the corporate memory and the decision outcome in past decisions. This function will require an a decision value. The function provides only indicative results. 
  • Predict & Save
      • Saves the cuent decision record to the database. The system will run and associate a decision prediction value, if a decision is provided.
  • Export Record
    • Exports the contents of the record (all tabs) to an html file on disk
  • Reset
      • Will reset all form values.
  • Help
      • Will display a help url on the Decide & Predict
References (Tab)

In this tab you will be able to provide references to other systems (Relationship manager, Marketing group, CRM, Client Id, Client group etc..) so that you will be able to run laters statistics on decisions and decision outputs.

Note the Status field, where you will be able to provide the status of your decision. The values in this field are defined by the application administrator.

Comments (Tab)

In this tab you will be able to add your comments on the decision. The Teaser text will be used as a header to your comment thread. Other users will be able to add comment threads to the decision.

ArrowCheck (Tab)

In this tab you will be able to document your decision. The interface is composed by Section and Folder tabs (under each section). Every tab may be given a name and in every Folder tab you can attach a documentation tool (~ 25 tools). The complete documentation produced in this tab along with the information provided in: Decision Attributes, References and Comments can be exported in a structured document (Export Record icon)


Manage & Document

In this tab you will be able to manage, document and comment existing decision records. You can filter the records you want to display and update the highlighted record. Note that depending on your rights and if a decision definitely applied to a case, you may not be able to update the decision basic elements or documentation but you can always add comment threads. As you can see below, the Decision tab displays the list of decision records with the contents of the highlighted record to be detailed in the other tabs. Note that the ribbon icons will depend on the tab you are working with. E.g. When in the Decision tab, you will be able to to filter the records through the Data Filter Builder or move around the records using the browser icons.

Decisions (Tab)

All decisions will be displayed in this tab (in a datagrid) in a descending order (last recrord first), with the ribbon icons that will allow you to browser the decision database and work with a specific record.

Let’s take a look to the ribbon icons and see how you can browse the list of records. Note that the system will display the records in pages, as defined in the Data Filter Builder query, getting all the records by default.


Let’s see how the hierarchy of the records and how they are displayed.

  1. Database Records
    This is the total number of database records. If database has many records (e.g. >= 1mio), the system will segment them in pages.
  2. Data Filter Builder Query Records
    This is the records retrieved from the database applying some query criteria. In this case the number of records will be smaller than the total number of database records.
  3. Page Records
    The query records are divided in pages. Every page will display by default 500 recs and the user will be able to decrease or increase this number.
  4. Total Number of Pages
    Once the page records defined, the system will divide the records per page by the total number of records per query to show the total number of pages.

  1. Page Records (dropdown).
    1. In this dropdown you will be able to define the number of recs you want to see by page. Note that although the visible are of the page may show only ~20 recs, you can browse vertically the page to see more records.
  2. Page Number/Number Of Pages
    1. In the Page Number field you can see the number of the current page and the total number of pages.
      You can move to a specific page by entering the page number and press Enter.
  3. Page Browsing Buttons
    1. Moves from page to page by pressing a single button without entering the page number.
  4. Page Records/Checked Records/Page Records Selected
    1. These are record indicators.
  5. Reference Fields
    1. Field selection tool
  6. Data Filter Builder
    1. Query tool
  7. Save Session To File
    1. Allows the saving of the session parameters to a local file for a later use.
  8. Save Default Configuration
    1. Saves the current session as the default session parameters
  9. CSV Recs Export
    1. Exports the page records to a csv local file on disk
  10. Export Record
    1. Exports the current highlighted record to an HTML file on disk.

Decide & Predict. References


In this tab, you will be able to add any references to your organization’s systems or references to entities, like customers, relationship managers, accounts etc.. These references will be very helpful later for your statistics.

The status field shows the status of the decision and can be used to show at which stage the decision is: e.g. Processing, Waiting Input, Escalated etc.. The status options are defined by the application Administrator.

Add New Decision


This is the New Decision form where you will enter the new decision data. The form is divided in different areas each one composed by different application elements:

Tabs

  1. Decision Attributes
    1. Decision. Contains the decision selection (severity level). Depending on your role, you may not be able to apply a Decision if the application is configured for a 4-Eyes Decision. A 4-Eyes Decision must validated by a different user or a supervisor.
    2. Decision Attributes. The decision characteristics (risk levels). Green highlighted. The decision attributes are configured and managed by the application Administrator. When a decision is taken, the user should provide for each decision characteristic the level of risk taken. The risk weight value (in parenthesis) shows the weight of the characteristic/attribute on the decision.
    3. Risk Weight. The risk weight per decision attribute (read only). Blue-highlighted. The risk weight values  are configured and managed by the application Administrator. They show the weight of the characteristic/attribute on the decision.
    4. Weighted Risk. The calculated decision weighted risk (read only). Red-highlighted. This value is re-calculate whenever a decision attribute or a risk weight value is modified.
  2. References
    1. In this tab you should provide the references internal to your organization. These are usually references to your internal IT system, project numbers, customer ids, etc.. Please note that these references will help you with the decision and risk analysis.
  3. Comments
    1. The user free comments on the decision. ArrowMiner comments are structured in threads that can not be edited once appended.
  4. ArrowCheck
    1. The decision documentation structured in Sections and Folders tabs. Each Section tab can contain one or more Folders and each folder may contain a Documentation Function (~20 available functions). All documentation Sections and Folder may be individually named and form a structured document.
Ribbon Icons

    1. Expand/Collapse. Focuses on the working tab.
    2. Decision Prediction. Runs a Decision Prediction based on previous decisions on similar weighted risk.
    3. Decision Prediction Based On Outcome. Runs a Decision Prediction based on previous decisions outcomes on similar weighted risk. This function will require an Estimated/Wanted Outcome.
    4. Outcome Prediction. Runs an Outcome Prediction based on previous outcomes on similar weighted risk.
    5. Predict & Save. Will run prediction and save the current record to the database:
    6. Export Record. Exports the record contents to an HTML file.
    7. Reset. Re-initializes the form.
    8. Help. Displays the help screen.
Questions & Answers

  • I do not see the Decision field but the 4Eyes Decision field. Why?
    • Depending on the application configuration and your role a 4-eyes decision may be required before this is validated with a final Decision.
  • Why the Decision Prediction shows a Severity Level and not a specific decision?
    • Decisions are grouped in Severity levels and options within severity levels. Decision Predictions will be always provided on severity levels.
  • Should I better consider a Decision Prediction based on previous decisions or outcomes?
    • Good decisions may produce good outcomes and vice versa. You should run both prediction and analyze the risk taken before you decide.
  • Can I modify the risk weight value for a specific decision?
    • You can do this, only if your role is >= Supervisor and if the application administrator has allowed this feature.
  • Can I save temporarily a decision and update it later?
    • Yes you can. You can save temporarily a decision (without applying a decision) and update it later.
  • Can I apply definitely a decision to a case and update it later?
    • No, you cannot. Once a decision is applied the decision cannot be updated. Only comment threads can be appended.
  • What is the Prediction Confidence %?
    • This is the number of decisions that have been identical to the predictions
  • How Can I know the Application and the Module I am working with?
    • You can always see this on the ArrowMiner header as shown below.
  • What are the Comment Threads and the Comment Teaser?
    • Comments may be added at any time by any allowed user to a decision without any length limitation.
    • Whenever a comment thread is added, the user should add a comment teaser/header.
    • All comment threads  are  shown in a date descending order with a single line with the name of the user, datetime of the thread and the teaser provided by the user (or the system). You can click on any comment thread teaser to expand the comment contents.
  • How can I print a decision record?
    • You can click on the Record Export icon and get an HTML file of the record.

Average Prediction Deviation


This is the distance between the Decision and the Prediction values (+/-). The value in this field should be compared to the Prediction Confidence. Best values are: Prediction Confidence: 100%, Average Prediction Deviation: 0.

0 value means: Decision = Prediction.

+ values mean: Prediction values are higher (severity).

values mean: Prediction values are lower (severity)