There should be approaches to automate repetitive, tedious, data-related work. Automation improves efficiency and effectiveness. Using scripts for automated continuous monitoring to detect exceptional data in transactions is one such approach.

How to create and use scripts to automate continuous monitoring?

• Definition of a script
• Logic and processes for creating a script
• Automated continuous monitoring via using a script
• The benefits of using a script to automate continuous monitoring
• Considerations for using a script to automate continuous monitoring

1. Definition of Script

A script is a computer language that contains multiple commands in a file using syntax. When opened by an appropriate scripting engine, the commands in the script are executed automatically without compiling. Scripts can be used to automate processes. The script mentioned in this article can also be understood as a simple input and output system.

The input is the source data from other systems or files

The output is the result of automatic analysis and calculation of the script.

2. Logic and Process for Creating a Script

• Understand the source data and data fields that need to be continuously monitored.
• Determine what kinds of data are to be imported into the script. The source data can usually be imported directly from the source databases through APIs.
• Other parameters and file templates for writing the script are placed in a fixed storage address.
• These data are automatically imported when the script is run by using the import command in the script.
• Determine the following basics:

– Which data needs monitoring and set thresholds for filtering out exceptional data?

– Frequency of monitoring (daily, weekly, monthly, quarterly, etc.).

– How the monitoring results are released (stored in an authorized shared server directory, or sent by mail, etc.).

– The recipients and reviewers of the monitoring results.

• Based on data monitoring scenarios, evaluate the different calculations and analysis methods to be used. Before writing the script, document the calculations and analysis methods.
• The export template of monitoring results is compiled and stored in a fixed storage address.
• Define the storage path of monitoring results or the email recipients’ addresses.
• After completing the above steps, use data analytics software to create a script by using syntaxes and commands.
• After the script is written, run the script and check the monitoring results. If there are errors, amend the script accordingly until it is accurate.
• According to the monitoring frequency, add a continuous monitoring start time to the script, and the script will automatically run at each start time.

In short, the logic of creating a script can be described as “begin with the end in mind”. According to each step of the analytics, the script is written by using computer syntaxes and commands to achieve the goal of continuously and automatically exporting the results.

3. Automatically Run Scripts

According to the defined automatic monitoring frequency, the scripts will automatically run. The monitoring results will be automatically sent to the preset authorized server directory. Each recipient can log in to the server to view the results or the monitoring results can be automatically sent to the email addresses of each recipient.

The reviewer of the monitoring results examines the monitoring results and confirms the followings:

• The exceptions in the monitoring results are indeed deviations. The deviations should be corrected and improved accordingly.
• If the exceptions are false alarms, which should be fed back to the person in charge of the script, and the script will be corrected accordingly.

4. The Benefits of Using Scripts to Automate Continuous Monitoring

• It solves the problem of bulk data analysis that is difficult to be analyzed manually.
• Human errors are avoided.
• Save a lot of time for data calculation and analysis.
• Timely notice the deviations in the transaction data and correct them in time to avoid further losses.
• Timely attention to potential fraud in transactions, and timely investigation, to avoid the occurrence of greater losses.

5. Considerations for Using scripts to Automate Continuous Monitoring

• Writing scripts using specific computer syntaxes and commands is relatively simple. It is more important to have a good understanding of the logic between the source data and monitoring results.
• Exceptional data need to be checked by the reviewer. Automatic continuous monitoring is useless without a manual review of monitoring results. The script cannot replace the manual review of exceptions.
• After noticing false alarms in monitoring reports, modify the script in time to eliminate these false alarms.
• After multiple runs of the script the following matters need to be considered while reviewing the exceptional data to improve its efficiency and effectiveness:

a. The possibility and necessity of modifying the threshold for exceptional data.

b. Adjust the frequency at which the script runs to ensure that monitoring results can be reviewed in a timely manner. If the reviewer is not offered enough time to review, possibly the report may remain unnoticed or unanalyzed.

For kinds of exceptional transaction data that can be continuously and automatically monitored by scripts, watch out for our next blog.