Input and output data

The Adverse Media Monitoring AI Agent supports a variety of ways for seamless integration into each financial institution’s ecosystem, such as client lifecycle management (CLM) systems, news sources, and case management archives, including through RPA and API connections.

Search input

The Adverse Media Monitoring search is triggered by specifying a name via the Ad-hoc search in Workspace or a batch upload of names set up and scheduled via Control Tower. To conduct a search, a name is required.

To configure a batch process, WorkFusion can be integrated with the bank’s internal systems to mimic how analysts are currently retrieving names that need to be monitored for adverse media events.

Examples include:

• Navigating alert queues in a transaction monitoring system and conducting searches on a non-customer in a potentially unusual transaction.
• Identifying customers who meet certain criteria within a CLM system. For example, you have a customer rated as high-risk, and yet the most recent Know-Your-Customer refresh is greater than one year.
• New applications created within a customer onboarding system or business line product extension system (such as a new lending application).
• A CSV file sent to a shared email inbox or uploaded to a shared folder.

You can also trigger a search via REST API that passes all required inputs as parameters and then receives the results from WorkFusion.

CSV input

The data input format for the Evan: Adverse Media Monitoring File Ingestion Business Process is a CSV file with simple fields and one or more rows with entity names to search.

Below is an example of the CSV file with possible content. You can also download the template when configuring the External source as the AMM input.

Input data schema

Only the search_request field is obligatory. All other fields are optional and are used to increase the relevancy of search results.

Field	Description	Type	Required	Example
search_request	Entity name. It is a person or a company subject to Adverse Media Monitoring.	String	YES	Danske Bank John Doe
entity_type	Entity type. Currently, two possible values are supported (non-case-sensitive): Company Individual	String	NO	INDIVIDUAL COMPANY
residence_operating_country	Two-letter country code. Supported format: ISO 3166-1 alpha-2. See the full list here.	String	NO	US (United States) / CA (Canada) / AU (Australia)
subdivision	Two-letter subdivision code according to ISO 3166-2.	String	NO	US-FL (Florida state of the USA) US-CA (California state of the USA) CA-QC (Quebec province of Canada) CA-YT (Yukon territory of Canada)
citizenship_incorporation_country	Two-letter country code. Supported format: ISO 3166-1 alpha-2. See the full list here.	String	NO	US (United States) / CA (Canada) / AU (Australia)
city	Full city’s name. Used only for Thomson Reuters CLEAR Investigations.	String	NO	New York
year_of_birth	Year of birth for a single screened person. Not applicable when entity_type is Company.	Integer	NO	1980
date_of_birth	Date of birth for a single screened person. Not applicable when entity_type is Company. List of accepted date formats: MM/dd/yyyy yyyy/MM/dd MM-dd-yyyy yyyy-MM-dd dd MMM yyyy dd MMMM yyyy The date format is selected in the course of configuration.	Integer	NO	02/25/1980 1980/02/25 02-25-1980 1980-02-25 25 Feb 1980 25 February 1980
additional_information	Additional information about the screened entity.	String	NO	John Doe has a brother Edward Doe
add_to_ongoing_monitor	The field is used only when Thomson Reuters is configured as the data provider. If true, the corresponding entity will be automatically added to the ongoing monitoring scope.	Boolean	NO	true/ false
_sys_mt_hit_assigned_user_groups	The name of the user group assigned to manage the manual review task in Workspace.	String	NO	Reviewers
article_data	The field is used only when the CSV file is selected as the input data source. It represents stringified JSON with an array of article data (links and additional information). For the full schema, see the Article data item table.	String	NO	[{\"link\":\"https://www.politico.com/news/2024/08/27/nancy-pelosi-jan-6-fury-00176529\",\"additional_information\":\"article level information\"}, {\"link\":\"https://www.bbc.com/news/articles/czdpq4en1vvo\"}]
entity_id	A unique ID of the screened entity used to re-apply decisions on the content previously reviewed by an analyst. Free text, in any format.	String	NO	E1234567
batch_id	A unique ID to group multiple search entities into combined reports. To be sent together with the batch_size field.	String	NO	Unique free text in any format. See some examples below: BI123 Batch01 Group01
batch_size	The number of investigations in the batch. To be sent together with the batch_id field.	Integer	NO	15
external_provider_name	The name of the provider used for collecting all articles.	String	NO	GOOGLE
external_id	A unique ID to identify the search request. Unique free text, in any format.	String	NO	R1234567
cost_center_code	Code of the Cost Center team for billing transactions.	String	NO	R1234567
cost_center_name	Name of the Cost Center team for billing transactions.	String	NO	R1234567
sla_date	Service Level Agreement (SLA) date for filtering in Manual Tasks. The date format is selected in the course of configuration.	String	NO

Article data item fields

Field	Description	Type	Required	Example
link	Link to the website from which an article is to be downloaded.	String	YES if content is null.	https://example.com/path/to/article
content	Article content. If both link and content are provided, content is overwritten with the content from link, provided the latter is successfully downloaded.	String	YES if link is null.	A large text that represents the article content in HTML or a plain text format.
title	Article title.	String	NO	Unlocking the Secrets of Everyday Life: How Small Habits Shape Big Success
summary	Article summary.	String	NO	In a world driven by goals and aspirations, the importance of daily habits often goes unnoticed. Yet, research suggests that small, consistent actions can lead to transformative results.
additional_information	Plain text or part of HTML containing additional information that can be helpful during the article review.	String	NO	Industry: Technology and Software Development Founded: 2015 Headquarters: San Francisco, California, USA
external_provider_name	The name of the provider used for collecting the article.	String	NO	Ultimate Article Searcher
investigation_status	If an article is initially reviewed with Evelyn NSS, you can pass the NSS decision and it will be considered by Evan AMM.	String	NO	NEEDS_INVESTIGATION FALSE_POSITIVE TRUE_POSITIVE
external_review_comment	If an article is initially reviewed with Evelyn NSS, you can pass the NSS comment and it will be considered by Evan AMM.	String	NO	This article was reviewed by Evelyn NSS as False Positive.

Search output

Once the entity search is completed, the results are displayed in a risk-prioritized manner as described above. The AI Agent user then reviews the results.

Once the user completes the review, the AI Agent generates a PDF report audit trail of the identified news, including the confidence score per article, the overall decision, and other data.

By default, the audit trail report is stored inside the WorkFusion platform. It is possible to configure this report to be pushed directly into organization’s internal applications, for example:

• Save into a case management or transaction monitoring system
• Store in a shared folder
• Use a custom connector for an in-house application

Output data schema objects

Investigation data schema

Field	Description	Type	Example
uuid	A unique identifier for the investigation	String	123e4567-e89b-12d3-a456-426614174000
runUuid	A unique identifier for the execution run	String	123e4567-e89b-12d3-a456-426614174001
batchId	A unique ID to group multiple entities into a combined report	String	B1234567
entityId	A unique ID for identifying entity	String	E1234567
externalId	A unique ID to identify the search request	String	R1234567
tenantId	A unique identifier for the tenant	String	tenant-123
variationId	Variation identifier	Integer	1
version	Version of the AMM AI Agent	String	4.0.2
status	Current status of the investigation	String	STARTED / IN_PROGRESS / READY_FOR_REVIEW / REVIEWED / COMPLETED / FAILED / MT_EXPIRED
inputStartTime	Timestamp when the input processing started	DateTime	S2024-03-18T10:15:30Z
inputEndTime	Timestamp when the input processing ended	DateTime	2024-03-18T10:16:30Z
processingStartTime	Timestamp when the main processing started	DateTime	2024-03-18T10:16:31Z
processingEndTime	Timestamp when the main processing ended	DateTime	2024-03-18T10:20:30Z
reviewStartTime	Timestamp when the review started	DateTime	2024-03-18T10:20:31Z
reviewEndTime	Timestamp when the review ended	DateTime	2024-03-18T10:25:30Z
outputStartTime	Timestamp when the output processing started	DateTime	2024-03-18T10:25:31Z
outputEndTime	Timestamp when the output processing ended	DateTime	2024-03-18T10:26:30Z
originalSearchRequest	Original search entity name as provided	String	John Doe
searchRequest	Processed search entity name	String	John Doe
entityType	Type of the search entity	String	COMPANY / INDIVIDUAL
city	Full city name	String	New York
residenceOperatingCountry	Two-letter country code in the ISO 3166-1 alpha-2 format	String	US
citizenshipIncorporationCountry	Two-letter country code in the ISO 3166-1 alpha-2 format	String	CA
subdivision	ISO 3166-2 code of the subdivision	String	US-FL
yearOfBirthIncorporation	Year of birth for an individual or year of incorporation for a company	Integer	1980
dateOfBirthIncorporation	Date of birth for individual or incorporation for company	Date	1980-05-15
htmlReportLink	Link to the HTML report	String	https://example.com/reports/123.html
pdfReportLink	Link to the PDF report	String	https://example.com/reports/123.pdf
investigationStatus	The resulting status of investigation	String	FALSE_POSITIVE / TRUE_POSITIVE / NEEDS_INVESTIGATION / NO_RESULTS_FOUND
restExecution	Indicates if the execution is via REST API	Boolean	true / false
assignee	User assigned to the investigation	String	john.smith@example.com
assignedUserGroups	The name of the user group to which a manual review task is to be assigned	String	reviewers
additionalInformation	Additional information of the screened entity	String	John Doe has brother Edward Doe
addToOngoingMonitor	Indicates if an entity should be added to ongoing monitoring	Boolean	true / false
costCenterCode	Code of the Cost Center team for billing transactions	String	CC12345
costCenterName	Name of the Cost Center team for billing transactions	String	Risk Department
batchSize	The number of investigations in the batch	Integer	15
uiConfiguration	UI configuration settings	String	{"dateFormat": "MM/dd/yyyy"}
providers	Array of providers used in the investigation	Array	[{"provider": "GOOGLE", "active": true}]
articles	Array of articles found during the investigation	Array	[{"title": "Example Article", "status": "PROCESSED"}]
articleCount	The number of articles found	Integer	25
manualTaskRequired	Indicates if a Manual Task is required	Boolean	true / false
manualTask	Manual Task details	Object	{"status": "NEW", "assignee": "john.smith@example.com"}
dateOfBirthMatchingThreshold	Threshold value for matching the date of birth	Number	0.85
mlModelId	Identifier for the used machine learning model	String	model-123
manualTaskEnabled	Indicates if Manual Tasks are enabled	Boolean	true / false
skipMtIfAllAreFalsePositive	Skip the Manual Task if all articles are false positives	Boolean	true / false
skipMtIfAllArePreviouslyReviewed	Skip the Manual Task if all articles were previously reviewed	Boolean	true / false
reportType	The report type to generate	String	HTML / PDF
businessProcessLanguage	Language for business processing	String	EN / ES
dispositionAll	Indicates if all articles should be dispositioned	Boolean	true / false
routeUserToNextTask	Indicates if a user should be routed to the next task	Boolean	true / false
identifyDuplicateArticles	Indicates if duplicate articles should be identified	Boolean	true / false
duplicateThreshold	Threshold for identifying duplicates	Integer	90
generateReportIfManualTaskExpires	Indicates whether a report should be generated even if a Manual Task expires	Boolean	true / false
useLlmModel	Indicates if an LLM model should be used	Boolean	true / false

Article schema

Field	Description	Type	Example
uuid	A unique identifier for the article	String	123e4567-e89b-12d3-a456-426614174002
newsProvider	Provider of the news article	String	GOOGLE / FACTIVA_HEADLINES / LEXIS_NEXIS_LP_MEDIA / THOMSON_REUTERS_CLEAR / WORLD_CHECK_ONE / CSV
duplicateArticleUUID	UUID of the duplicate article if found	String	123e4567-e89b-12d3-a456-426614174003
title	Article title	String	Example News Article
status	Processing status of the article	String	PROCESSED / UNPROCESSABLE / EXCLUDED
publishingDate	Date when the article was published	Date	2024-02-15
summary	Article summary	String	Brief summary of article content
author	Article author	String	Jane Smith
language	Language of the article	String	EN / ES / UNKNOWN
wordCount	Number of words in the article	Integer	850
source	Source of the article	String	The New York Times
originalArticleLink	Original link to the article	String	https://example.com/article/12345
downloadedArticleLink	Link to the downloaded article	String	https://storage.example.com/12345.html
parsedArticleLink	Link to the parsed article	String	https://storage.example.com/12345.html
taggedArticleLink	Link to the tagged article	String	https://storage.example.com/12345.html
investigationStatus	Status of the investigation for the article	String	FALSE_POSITIVE / TRUE_POSITIVE / NEEDS_INVESTIGATION
adjudicationReason	Reason for adjudication	String	NON_ADVERSE / NON_FOCAL / YOB_MISMATCH / ENTITY_WAS_NOT_FOUND / DUPLICATE / ARTICLE_PARSING / UNSUPPORTED_LANGUAGE / UNSUPPORTED_FORMAT / BEHIND_PAYWALL / HUGE_ARTICLE / POOR_ARTICLE / EMPTY_ARTICLE / DOWNLOAD_ARTICLE_EXCEPTION / WEBSITE_IGNORE / ARTICLE_ON_SANCTIONS_SITE / NSS_DECISION / LLM_RESPONSE_FALSE_POSITIVE / LLM_RESPONSE_NEEDS_INVESTIGATION
mlScore	Machine learning score	Number	0.87
taggedTitle	Tagged version of the title	String	Example tagged title
taggedSummary	Tagged version of the summary	String	Example tagged summary
taggedHighRiskCountries	Tagged high-risk countries	String	US / UK /RU
taggedKeywords	Tagged keywords	String	fraud / money laundering
taggedPoliticallyExposedPersons	Tagged politically exposed persons	String	John Smith / Jane Doe
taggedEntities	Tagged entities	String	ABC Corp / XYZ Ltd
errorCode	Error code if processing failed	String	ARTICLE_DOWNLOAD / ARTICLE_CONTENT_EMPTY / UNSUPPORTED_LANGUAGE
historyId	History identifier	String	histqwdwqdund234dwqdwqd21und2345hqwe12d2345
additionalInformation	Additional information about the article	String	Contains references to subsidiaries
reviewComment	Comment from the review	String	Not relevant to the search entity
extractedYearOfBirthIncorporation	Extracted year of birth or incorporation	Integer	1981
providerSearchUrl	URL used for the provider search	String	https://search.example.com?q=John+Doe
articleReview	Article review details	Object	{"workerFullName": "John Reviewer", "investigationStatus": "FALSE_POSITIVE"}
previousReview	Previous review details	Object	{"workerFullName": "Jane Reviewer", "investigationStatus": "NEEDS_INVESTIGATION"}
contentParsedBy	Tool used to parse the content	String	POSTLIGHT / TRAFILATURA
authorParsedBy	Tool used to parse the author	String	POSTLIGHT / TRAFILATURA
publishingDateParsedBy	Tool used to parse the publishing date	String	POSTLIGHT / TRAFILATURA
titleParsedBy	Tool used to parse the title	String	POSTLIGHT / TRAFILATURA
externalProviderName	Name of the external provider	String	GOOGLE
downloadMode	Mode used for downloading	String	STANDARD / SCRAPE_DO_STANDARD / SCRAPE_DO_BROWSER_RENDERING

Article review schema

Field	Description	Type	Example
workerUUID	UUID of the worker who performed the review	String	123e4567-e89b-12d3-a456-426614174004
workerFullName	Full name of the worker	String	John Smith
workerEmail	Email of the worker	String	john.smith@example.com
investigationStatus	Status determined by the investigation	String	FALSE_POSITIVE / TRUE_POSITIVE / NEEDS_INVESTIGATION
comment	Review comment	String	Not relevant to the entity being searched
reviewTime	Timestamp of the review	DateTime	2024-03-18T14:30:45Z

External article data schema

Field	Description	Type	Example
articleLink	Link to the article	String	https://example.com/article/12345
investigationStatus	Status determined by the investigation	String	FALSE_POSITIVE / TRUE_POSITIVE / NEEDS_INVESTIGATION
externalReviewComment	Comment from an external review	String	This article was reviewed by Evelyn NSS as False Positive
externalProviderName	Name of the external provider	String	GOOGLE
additionalInformation	Additional information about the article	String	Industry: Technology and Software Development
title	Article title	String	Example News Article
summary	Article summary	String	Brief summary of the article content
content	Full article content	String	A large text that represents the article content in HTML or a plain text format

Manual Task schema

Field	Description	Type	Example
workerUUID	UUID of the worker assigned to the task	String	123e4567-e89b-12d3-a456-426614174005
workerFullName	Full name of the worker	String	John Smith
workerEmail	Email of the worker	String	john.smith@example.com
status	Status of the Manual Task	String	NEW / IN_PROGRESS / SUBMITTED / EXPIRED
investigationStatus	Status determined by the investigation	String	FALSE_POSITIVE / TRUE_POSITIVE / NEEDS_INVESTIGATION
reviewComment	Comment from the review	String	No relevant findings detected
reviewTime	Timestamp of the review	DateTime	2024-03-18T14:30:45Z
processingPeriodSeconds	Time spent processing in seconds	Integer	300
assignee	User assigned to the task	String	john.smith@example.com
assignedUserGroups	User groups assigned to the task	String	reviewers
selectedForAutoQc	Indicates if the task is selected for automatic quality control	Boolean	true / false

Provider schema

Field	Description	Type	Example
provider	Provider name	String	GOOGLE / FACTIVA_HEADLINES / LEXIS_NEXIS_LP_MEDIA / THOMSON_REUTERS_CLEAR / WORLD_CHECK_ONE / CSV
externalProviderName	Name of the external provider	String	GOOGLE
externalArticlesData	Array of external article data	Array	[{"articleLink": "https://example.com/article/123", "title": "Example Article"}]
keywords	Array of keyword data	Array	[{"keywords": "fraud,laundering", "enableKeywordThesaurus": true}]
active	Indicates if the provider is active	Boolean	true / false
highRiskCountries	High-risk countries for the provider	String	US / UK / RU
daysSearchPeriod	The number of days to search	Integer	90
searchArticleCount	The number of articles to search	Integer	100
searchLanguage	Language to search in	String	EN / ES / ANY / EN_ES
excludedFileTypes	File types to exclude	String	pdf / doc / txt
dynamicLocationSearchEnabled	Enables dynamic location search	Boolean	true / false
partialNameSearchEnabled	Enables partial name search	Boolean	true / false
ignoreDeadLinks	Indicates if dead links should be ignored	Boolean	true / false
ignoreRedirects	Indicates if redirects should be ignored	Boolean	true / false

Keyword data schema

Field	Description	Type	Example
keywords	Keywords to search for	String	fraud / laundering / bribery
enableKeywordThesaurus	Enables the keyword thesaurus	Boolean	true / false