About Sherlock

What is Sherlock?

It is a service of Masivian's security suite. It uses artificial intelligence to prevent fraud based on the analysis of the historical behavior of the telephone lines of our customer's users. It is composed of several indicators that complement each other to facilitate fraud detection and prevention.

Indicators

Anomaly

Detection of usual behavior change of transactional and non-transactional messages associated to a telephone number.

Examples:

• Detect an unusual number of transactions associated with a line (spoofing).
• Detect if a number has recently stopped receiving messages (number behaviour change).

Distrust

Evaluation of the transactions associated with the telephone number compared to the behavior of the lines in the country. A low distrust represents that the number has a normal transactional behavior compared to the rest of the lines in the country.

Examples:

• Prevent possible fraud attempts with new or low reputation numbers.
• Detect numbers that do not have a history of usage and message reception.

Portability

Indicates if a number has changed its mobile operator in the last 7 days.

Example:

• Indenty theft of a person with mobile operators to obtain their telephone number.

Historical portability

Indicates the historical portability associated a telephone number in Colombia or Peru.

Example:

• If a phone number have portability to another operators, it show the historical since original operator until actual operator.

Pair matching

Indicates the information of cards/accounts that is associated from a telephone number.

Example:

• If a phone number have a card associated with the last 4 numbers have given, it'll shows true.
• If a phone number have a account associated with the last 4 numbers have given, it'll shows true.

SIM change

Indicates whether a telephone number has a SIM change in a period determined by the customer or not. It is more specific than the portability service because it includes the detection of changes in the SIM card associated with the number.

Example:

• Identity theft to get the phone number from the mobile operators.

Phone active age

Estimation of the time from which a line has been used. It detects and evaluates periods of inactivity, which allows to calculate the time of use of the current user.

Example:

• If a phone number was disabled and then was assigned to another person, Sherlock recognizes the age of the line since the last activation.
• Prevent possible fraud attempts with new or recently reactivated numbers.

Categorization of Comms

Analyzes the total amount of SMS sent to a customer and calculates the percentage of messages sent from each of the categories and segments of the industry for each of the lines consulted.

Example:

• A line that receives a high percentage of notifications from a collections department may not be appropriate for credit approval.

Third party reporting (Line Status Manager)

It allows customers to report a fraudulent event to a specific line so that other associated entities can see this report and determine the procedure they are performing.

Example:

• A line that performed a fraud can be reported by the affected customer. The other areas and entities that consume Sherlock can see this report and thus prevent fraud.

Customer use cases

• Validate the opening of accounts and obtaining new products or services.
• Validate new bank accounts and its transactions.
• Decision making to enable OTP transactions and remittances.
• Generation of alarms once a suspicious line has been detected for management by the client.

Stakeholders

• Entities belonging to different industry sectors, mainly from the banking sector.
• Internal products Masiv as a Transactional Hub

Commercial staff in charge

Jose Maria Serrano