Uncategorized

Users can use the ‘Add section‘ feature in Document type to create repeating fields and tables, allowing them to group these elements using header, start, and end flags. Within each section, users also have the flexibility to define individual fields and tables.

Users must have any one of the following policies to add section and group in Document type

• Administrator Policy
• Creator Policy

.

Step 1: Create a New Document Type

1. Head to the Administration module, choose Document Type and select Create New.
   
   
2. In the Create new Document type page, enter the document type Name, Description and add an Image.
   
   
   

Step 2: Add section

1. In the Create new Document type page, select  Add section.
   
2. The section pane that appears, enter Section header name.
   
   
   
3. You can add fields/tables in this section.
4. To add a new field against the section, continue with Step 3. 
5. To add a new table against the section, continue with Step 5.

Step 3: Define field in a section

You can add a field inside the section and also outside of the section. This information allows you to create a field inside the section. 

1. In the section pane,, Select Add field
   
2. In a field that appears, Enter the field Name, and select the appropriate data type from the drop down list. 

• Use Delete to delete the table. 
• Select to add additional fields. 
• Select to add tables in the section. For more information about add tables, see Step 5

Step 4:  Field settings 

The platform offers you the following settings for a field:

Expected Label Output:

These options allow you to set the expected output during the result against a field. In documents

1. In the Section pane, Click Settings against a field and choose any one of the following Expected Label Output options:
   • Required once: This option enables the section field to annotate multiple entities against a field and generate a single result. Use this option when a document contains the same value in multiple places in a page.  
   • Required multiple: This option enables the section field to annotate multiple entities against a field and also generate multiple results against a field. 

Selection Settings

These settings allow you to set the position of the field in a group. 

2. In the Section settings pane that appears, choose the following labels to highlight the label.
   • Header:  This label indicates a header of a group. Choose this label if you wish to annotate the header entities against a field.
     
   • Start:  This label indicates a field is the starting field of a group. Choose this label if you wish to annotate the starting entities against a field.
     
   • End: This label indicates a field is the ending field of a group. Choose this label if you wish to annotate the ending entities against a field. 

PII

• Select PII check box  to encrypt the field value. Use this option for the  PII (Personally Identifiable Information). This option ensures data security of the field information. 
  
• Click Submit 

Step 5: Add table in the section

You can add a table inside the section and also outside of the section. This information allows you to create a table inside the section. 

3. In the section pane, Select Add field.
   
4. In the table pane that appears, enter the table name and click Add field.
   
5. In the column header  field that appears, Enter the field Name, and select the appropriate data type from the drop down list.
   
   

• Select to add additional column headers. 

Step 6: Table settings

Step 4:  Field settings 

The platform offers you the following settings for a field:

Expected Label Output:

These options allow you to set the expected output during the result against a field. In documents

3. In the Section pane, click Settings against a field and choose any one of the following Expected Label Output options:
   • Required once: This option enables the section field to annotate multiple entities against a field and generate a single result. Use this option when a document contains the same value in multiple places in a page.  
   • Required multiple: This option enables the section field to annotate multiple entities against a field and also generate multiple results against a field. 

Selection Settings

These settings allow you to set the position of the field in a group. 

4. In the Section settings pane that appears, choose the following labels to highlight the label.
   • Header:  This label indicates a header of a group. Choose this label if you wish to annotate the header entities against a field.
     
   • Start:  This label indicates a field is the starting field of a group. Choose this label if you wish to annotate the starting entities against a field.
     
   • End: This label indicates a field is the ending field of a group. Choose this label if you wish to annotate the ending entities against a field. 

PII

Select PII check box  to encrypt the field value. Use this option for the  PII (Personally Identifiable Information). This option ensures data security of the field information.

Retraining is the process of updating an Asset with new data obtained from both new documents and existing documents, along with their annotations. As data distributions change and new patterns emerge, retraining helps the Asset to adapt to the variations in the document sets and maintain its predictive power. This process also helps the Asset to continue performing accurately and effectively over time. 

Step 4: Select documents

Step 5: Annotate

If the desired accuracy has been achieved, you can proceed to Publish the asset. For more information on how to publish the Asset, see the Publish the asset. 

Note: To identify the retrained Asset, check its version. For example, if the initial Asset version is 1.0, the retrained Asset’s version will be 2.0. You can avail both versions in the Asset Studio. 

Once the retrained Asset is published, you can download the API and its documentation. The API can be invoked independently or used within a specific use case. If you wish to consume this Asset via API, see Consume an Asset via API.

It is recommended to use URL aliases, if you wish to consume multiple versions of an Asset. It allows you to consume its different versions via a single API. For more information, see URL aliases.

You can also consume this Asset in the Asset Monitor module. For more information, see Consume an Asset via Create Transaction.

Users

Create a new user

1. Head to the Administration module and select User management.
   
   
   
   
2. In the Users tab, click Create User.
   
   
   

3. On the Create new user page that appears, enter the First name, Last Name, a unique User name and Email.
4. Select the appropriate predefined Policies from the drop down list and then click Submit. 

Note: You can assign multiple policies to a user.

Note: The platform offers predefined policies that are configured based on attributes. To customize these policies, contact our support team at mapsupport@intellectdesign.com. For more information about policies, see Policies  section.

Edit user account information

This section provides instructions on how to edit user account information on  the platform. 

Note: For SSO users, only the policies field can be edited.

1. Head to the Administration module and select User management.
   
   
   
   
2. In the Users tab, select the user whose information you wish to modify.
   
   
   
   
3. On the User account information page that appears, click Edit account.
   
   
   
   
4. Make the desired changes to the user’s information and click Submit.
   
   
   

Note: You can assign multiple policies to a user. 

Remove a user

This section provides instructions on how to delete an existing user from the platform.

Note: SSO users who no longer need access must be removed from the platform and the in-house SSO portal to prevent unauthorized access.

1. Head to the Administration module and select User management.
   
   
   
   
2. In the Users tab, select the user you wish to delete. 
   
   
   

3. On the User account information page that appears, click and select Delete.

Reset password

This section provides instructions on how to reset the password. 

Note: This feature is not applicable for users who sign in using Single Sign On (SSO).

1. Head to the Administration module and select User management.
   
   
   
   

2. In the Users tab, select the user whose password you wish to reset.
   
   
   
   
3. On the User account information page that appears, click Reset password. The respective user will then receive an email to reset their password.
   
   
   
   

Attribute based access control

Attribute Based Access Control (ABAC) is a method of controlling access for users based on their attributes (characteristics), the resource they are trying to access, and the environment in which the access is being requested. 

According to NIST, ABAC is defined as “an access control method where subject requests to perform operations on objects are granted or denied based on assigned attributes of the subject, assigned attributes of the object, environment conditions, and a set of policies that are specified in terms of those attributes and conditions.”

Access control methods protect your sensitive digital assets. The two most recommended access control methods today are role-based access control (RBAC) and attribute-based access control (ABAC). The main difference between RBAC and ABAC is the method of providing access. RBAC grants access based on roles, while ABAC grants access based on a wider range of factors, such as the user’s attributes, the object’s attributes, and the action being performed.

ABAC is more flexible and granular than RBAC, allowing for fine-grained control and dynamic decision-making. The Platform uses ABAC to control access to its resources. This means that users can only access the modules and permissions that they are authorized to access, based on their attributes, the resource they are trying to access, and the environment in which the access is being requested. The policies define which users have access to which modules and permissions. For more information about the policies, see Policies section.

Policies

A policy is the set of permissions that determines what actions a user can perform on the platform modules and its resources. For example, a policy can permit users to access one or more modules, and to some or all the actions associated with those modules. This approach allows users to only access the information and features that they need, which can improve security and usability.

Currently the platform offers seven predefined policies such as Creator, Annotator, Viewer, Manager, Reviewer, Administrator, Gen AI User and Gen AI Consumer.

Note: The platform also offers customized policies. Contact our support team at mapsupport@intellectdesign.com to create customized policies.

Predefined policies

The following table displays the predefined policies along with their descriptions:

Policy	Description
Creator	The creator policy allows users to have full access to the Asset Studio, Document Library, Asset Monitor and Human-in-the-Loop modules. Creators can create and manage document sets, build, deploy, consume, and monitor assets. They also have limited access to the Administration module to only define and manage document types.
Annotator	The annotator policy allows users to only annotate assets in the Asset Studio module. Annotators can also delete annotations, import and export annotations, and view the history of annotations.
Viewer	The viewer policy allows users to access the platform modules such as Asset Studio, Asset Monitor and Human-in-the-Loop. Viewers are read-only users and cannot create, edit, or delete assets.
Manager	The Manager policy allows users to access only the Asset Monitor module. Managers can monitor the assets, activate and deactivate assets, enable manual review, and set threshold limits.
Reviewer	The reviewer policy allows users to access only the Human-in-the-Loop module. Reviewers can review transactions, verify and correct the fields that have fallen below the threshold settings.
Administrator	The administrator policy allows users full access and permissions to all modules in the platform. Administrators have exclusive permissions to add and manage users and their permissions within the platform. They can also establish connectors for document upload.
Gen AI User	The Gen AI User policy allows users to access the Generative AI Assets, Knowledge Base and tools in the Asset Studio module. It helps the Users to create and manage Generative AI assets, Knowledge Bases and tools.
Gen AI Consumer	The Gen AI Consumer policy allows users to access only published agents, ensuring no access to other modules for improved security and control.

The following table shows the modules and the permitted actions that users can perform using each policy:

Modules	Permissions	Administrator	Creator	Viewer	Reviewer	Manager	Annotator	Gen AI User	Gen AI Consumer
Administration	Add, edit and remove users	✔
	Assign predefined policies	✔
	Create, edit and delete the document type(s)	✔	✔
	Add, configure and delete connectors	✔						✔	✔
	Add, edit and delete the URL aliases and Download the  API	✔
Document Library	Create, update and delete document set(s)	✔	✔					✔	✔
Asset Studio	View all Classic assets, Use cases and find Classic assets, Use cases	✔	✔	✔			✔
	View all Generative AI Assets							✔
	View only the Published Generative AI Assets								✔
	Create, add document type, delete document type, annotate, train,fine-tune,validate, select and deselect test data, review results, publish,retrain, download the Asset API, upload Asset image, edit Classic Asset name, and delete an Asset.	✔	✔
	Annotate documents,, import, export annotation and review annotation	✔	✔				✔
	Create, Design, Edit Use case with Classific AI Asset	✔	✔
	Create, Design, Edit, Publish Use case with both Classic and Generative AI Assets	✔
	Create, Design, Edit, Publish the Use case only with Generative Assets, Knowledge Base and tools.	✔						✔	✔
	Add, Edit, Delete the Generative AI Asset(s)	✔						✔	✔
	Create, Edit, Publish Delete the Knowledge Base	✔						✔	✔
	Create, Edit, Delete the Tools.	✔						✔	✔
Human-in-the-Loop	View the list of transactions flagged for manual review	✔	✔	✔	✔
	Review, verify and correct the fields that have fallen below threshold limits	✔	✔		✔
Asset Monitor	View performance metrics of operational assets, find an assets	✔	✔	✔		✔		✔	✔
	Activate and deactivate assets	✔	✔			✔		✔	✔
	Enable manual review and set threshold limits	✔	✔			✔
	Consume/Try an asset via Create Transaction	✔	✔			✔		✔	✔

Assign a predefined policy

New User

You can assign a predefined policy to a new user when creating their account. For more information about assigning a predefined policy to a new user, see Create a new user section. 

Existing User

To update the policy for an existing user, follow the instructions below:

Note: If you are an SSO user, you can only access the platform with the Viewer Policy during your initial login. Contact the Platform Administrator if you wish to modify your policy.

1. Head to the Administration module and select User management. 
   
   
   

2. In the Users tab, select the user whose policy you wish to update.
   
   
   
   
3. On the User account information page that appears, click Edit Account.
   
   
   

4. In the Policies field, assign the appropriate policies.
   
   
    

Note: You can assign multiple policies to a user.

Set accuracy threshold

1. In the Asset Monitor page, click next to the specific asset you want to set a threshold on and select Settings.
   
   
   
2. A window will appear.
   
   
   
   
3. Click on an asset and enable Manual Review.
4. Set the threshold limit for the accuracy of respective fields. This determines the minimum acceptable accuracy level  required for each field to meet quality standards.
5. Click on Next and ensure that the asset is activated. Then, click Submit.
   
   
   

Human-in-the-loop review

1. If the accuracy falls below the set threshold, the document will appear in the HITL (Human-in-the-Loop) queue.
   
   
   
   
2. In the Human-in-the-Loop page, select the document for review. 
   
   
   
   
3. By default, the Sort fields by need review toggle is enabled. This ensures that fields requiring review will appear at the top of the review section.
4. To review and/or update a field, select the field and draw a bounding box on the document or manually edit the field value. 
5. Once the field is reviewed, you can tab out and the Review needed icon disappears and is updated with the Reviewed icon. 
6. Once you have reviewed the fields, click Submit to be directed to the Human in Loop transaction listing page or click Submit and Move Next to review the next transaction in the queue.

Monitoring key metrics is essential for assessing the performance and effectiveness of a document AI Asset that is consumed as an API. With the Asset Monitor module, you can actively monitor real-time transactions, view various metrics and Asset activation status.

Monitor Asset

1. Head to the Asset Monitor module, you can visualize the Asset information.
   
   
   
   
   • Optional: Click the Failed notification to know the failed reasons. 
   • Optional: Click the Review notification to review the documents. 

Grid columns 

1. On the Asset Monitor page, click Grid column option.
   
   
   
   
2. In the Grid columns that appear, select the required metrics for monitoring.
   
   
   
   

Note: The column information may differ based on the Assets. 

View Asset Metrics

1. On the Asset Monitor module, click the respective asset for which  you wish to view the metrics details.
   
   
   
   

Note: If you are facing difficulty to find the Asset. See Find an Asset.

2. In the Overview page that appears, you can evaluate the operational efficiency and usage patterns of an Asset from the following metrics.
   
   
   

Classic AI Assets	Generative AI	Use cases
Total transactions – Total transactions represent the overall number of API calls made to the asset during the specified period.  Documents processed: Indicates the cumulative number of processed documents, showcasing the asset’s effectiveness in handling various document types. Pages processed:   Indicates the cumulative number of processed pages, showcasing the Asset’s effectiveness in handling various document types. Average Duration:  Provides the average duration required to process a transaction, giving an overview of the efficiency and performance of the asset. Transaction by status: Displays the distribution of transactions based on their status, allowing for easy identification of completed, in-progress, failed, or pending transactions. Graph: Average processing time per transaction, Confidence score per transaction	Total transactions – Total transactions represent the overall number of API calls made to the asset during the specified period. Input Tokens: Total number of initiated tokens Completion tokens: Total number of tokens generated by the model to complete a given prompt. Total number of completed tokens. Average Duration: Provides the average duration required to process a transaction, giving an overview of the efficiency and performance of the asset. Transaction by status: Displays the distribution of transactions based on their status, allowing for easy identification of completed, in-progress, failed, or pending transactions. Graph: Average processing time per transaction, completion token for transaction.	Total transactions – Total transactions represent the overall number of API calls made to the asset during the specified period. Input Tokens: Total number of initiated tokens Completion tokens: Total number of tokens generated by the model to complete a given prompt. Total number of completed tokens.  Average Duration: Provides the average duration required to process a transaction, giving an overview of the efficiency and performance of the asset. Transaction by status: Displays the distribution of transactions based on their status, allowing for easy identification of completed, in-progress, failed, or pending transactions.

4. Use Filter option to view the metrics for the required time frame to evaluate the operational efficiency and usage patterns of an Asset by analyzing the below metrics over a specific time period, ranging from the 24 hours to one year. Also you can customize the time frame.
   
   

Transaction

1. Head to the Transaction tab.
   
   
   
   
2. In the Transaction tab that appears, you can view the transaction information. 
   
   
   
   

Note: Use the Search Bar and Filter to find your transaction. If you wish to create a transaction from this page, see Consume an Asset via Create Transaction.  

3. You can view transaction details including the Transaction ID, Status, Input tokens (or) No. of documents, completion tokens (or) No. of pages, model cost, initiated on,  Duration for further analysis and monitoring. 

Access your account to start leveraging the features of the platform. This guide will walk you through the steps on how to log in.

Login through email

You will receive an email from the platform team with the URL, user id and temporary password which the user can use to login. Open your preferred web browser and enter the platform URL “https://uk-sjp.intellectai.com/purplefabric/” in the address bar and press Enter. 

1. On the login page, enter your registered username or email address and click on Continue.
   
   
   

2. Enter your password in the password field.  
   
   
   
3. Click on Login to initiate the login process.
4. If you have forgotten your password, you can reset it by clicking the Forgot password? link.
5. If you are still having trouble logging in, you can contact customer support for help.

Login through SSO

We also provide SSO capability. If you have opted for SSO integration and login, you would be able to login with your network ID on your inhouse portals.

Platform modules

1. Click on and you can view the list of modules in the platform.
   
   
   
   

• The Asset Studio module allows you to build and manage the below document AI models which we refer to as “Assets.”
  • Classic AI
    • Classifier Asset
    • Extractor Asset
  • Generative AI
    • Automation Agent
    • Conversational Agent
  • Use cases
    
• The Document Library module allows you to create and manage document sets required for training and testing the Assets.
• The Asset Monitor module allows you to monitor Assets in real time, enable human in the loop, activate/deactivate Assets and consume Assets.
• The Human in Loop module allows you to review transactions and correct fields that have fallen below the threshold settings.
• The Administration module allows you to define user management policies, establish connectors for document upload and administer the document types.