PowerPoint Presentation

[Audio] Good morning everyone. As you can tell by the title, today I will be presenting Product Tutorials as part of our overall plan to reduce Mean Time To Identify and Mean Time to Repair. To achieve this, we will be leveraging ELK + Kibana Elasticsearch to store data, Logstash and Filebeat for data processing, Kibana for visualization, and alerts for relevant users when an issue occurs as well as scheduled reports to be shared. Let's get started..

[Audio] Our presentation today focuses on Product Tutorials related to ATM problems that have an impact on the customer experience. Issues range from maintenance tasks like automatic closure, cash dispenser errors, and more. We have seen an average of 150-200 incidents per month, with 20% classified as fatal errors by our ESQ team and 80% as warnings or suspect errors. It is paramount for us to prioritize repeated problem identification in order to understand customer needs and concerns. Additionally, we strive to ensure customer complaints are handled in a timely manner, although we are seeing issue resolution times of more than 1 month at present. We trust today's presentation will assist in finding more effective ways to identify and resolve ATM issues so that customers can have a positive experience..

[Audio] Our proposed solution will provide you with ELK + Kibana Elasticsearch for faster and more cost-efficient data storage. Logstash and Filebeat will ensure data is stored quickly and efficiently. Kibana will give you enhanced visualization, and you will be notified when an issue arises so you can take care of it quickly. Scheduled reports will also be provided to guarantee the accuracy of data. We will be able to reduce your Mean Time To Identify (MTTI) to a matter of minutes or hours, instead of days, and your Mean Time To Repair (MTTR) to just a few days, instead of weeks or months. This proposed solution will offer various scenarios, shown on a dashboard with detailed information, with notifications given when an issue arises. Reports will be provided on a regular basis to ensure accuracy and transparency..

[Audio] This slide is an example of a system generated dashboard report and alerts that are sent out at regular intervals. This is advantageous as it allows users to keep track of their data and take necessary actions in a timely manner. Our ELK+Kibana stack allows us to view, analyze, and interact with our data and notifications in real time, thus decreasing Mean Time To Identify and Mean Time to Repair..

[Audio] A screenshot of a computer is displayed on the left and a graphical representation of the error count per error type distributed across a certain time period is shown on the right. The tabular representation underneath shows the distribution of errors across error types. ELK + Kibana Elasticsearch, Logstash, Filebeat, and Kibana are used to propose a solution that will reduce the mean time to identify and repair issues. This is meant to provide an efficient and valuable tool to identify and address issues occurring with data in an automated fashion..

[Audio] Error reporting and analyzing is an integral part of any system. Our product offers an easy and comprehensive way to report and analyze errors with Error Distribution and ATM ID wise Error Distribution. Error Distribution shows distribution of different types of errors over the total number of errors in the system, while ATM ID wise Error Distribution shows distribution of errors per ATM ID over the total number of errors in the system. With this feature, it becomes easier to identify and troubleshoot individual errors quickly and efficiently. This helps reduce Mean Time To Identify (MTTI) and Mean Time to Repair (MTTR)..

[Audio] Tag Cloud is a powerful tool which displays visual representation of the count of errors per type. The size of the text is utilized to show the amount of each type of error. This makes it possible to identify areas that need improvement in order to reduce the Mean Time To Identify and Mean Time To Repair..

[Audio] The graph and table provide a representation of the errors being monitored within ATMs. The graph displays a visual of the distribution of the error counts, while the table includes more detailed information on the errors and the ATM IDs. This data can be exported in a comma-separated values (CSV) format for further analysis..

[Audio] A detailed dashboard is displayed, allowing the user to sort and filter data by a range of criteria. An in-built messaging system is included, displaying log messages in a pop up window for easy access and comprehension of the data. Furthermore, all reports can be exported in .csv format, providing the user with the potential to access and analyse the data from their own device with great speed and ease..

[Audio] The image demonstrates detailed metrics of one of the scenarios, including figures that show the distribution of a number of keywords per ATM ID. This data is beneficial for understanding how our ELK + Kibana Elasticsearch solution can minimize Mean Time to Identify and Mean Time to Repair by providing us with the essential insight to take accurate action..

[Audio] The architecture our company used to reduce Mean Time To Identify (MTTI) and Mean Time to Repair (MTTR) is represented by this slide. It shows that ELK+ Elasticsearch is used for data storage, Logstash and Filebeat for data processing, with Kibana for visualization. Automated alerts are sent to relevant users when an issue arises. Additionally, scheduled reports are shared with all stakeholders. Through the use of this architecture, our company has achieved its aim of reducing MTTI and MTTR..

[Audio] Logging of events can be extremely useful for quicker identification of problems in applications. With the help of filebeat, capturing multiline logs is possible, allowing us to view events in between the start and end points of the transactions. As an example, if we take a look at a payment transaction, we can observe the Initiating Cardless Session event at 12:54:20, followed by the Forcedeposit event. Subsequently, the Session Ended event is registered at 12:56:49. Moreover, we can also observe Type I or Type II scenarios. Utilizing the ELK+Kibana stack allows for detection and visualization of these transactions and their related events, thus enabling us to identify problems with a more comprehensive view in a shorter span of time..

[Audio] We addressed the issue by making the necessary changes in Filebeat to ensure successful reading of all files. We changed the encoding to UTF-16le-bom, which allowed us to read all log files and identify errors based on keywords combinations found in the logs. Additionally, we developed a ‘Failure Handling Mechanism’ to notify the application owner if there were any logical errors at Logstash level. We also discovered that log files were not being overwritten and needed to be recreated at intervals of 6 hours. As a result, Filebeat identified them as new files and reingested the same data. Moving on, let's talk about how physical activity and exercise can reduce stress and increase energy levels..

[Audio] To guarantee the successful application of our solution, access to the data source server, the Logstash server and Kibana must be approved, and secure endpoints for both Logstash and Elasticsearch servers need to be established. Additionally, SSL/TLS certificates must be enabled for additional security, and access to both User Acceptance Testing (UAT) and production environments must be enabled. Furthermore, it is indispensable that all concerned stakeholders comprehend the project requirements, data sources and desired outcomes..

[Audio] "At OTWM Kiosk Technologies, we have a track record of excellence when it comes to providing high-quality solutions that enable businesses to streamline processes and maximize efficiency. Our Kiosk Technologies suite of products is designed to do just that and helps companies ensure that their data is organized, secure, and accessible. With Kiosk Technologies, businesses can access the benefits of ELK + Kibana Elasticsearch to store data, Logstash and Filebeat for data processing, Kibana for visualization, alerts for relevant users when an issue occurs, and scheduled reports to be shared; all of which help reduce Mean Time To Identify (MTTI) and Mean Time to Repair (MTTR). Thank you OTWM Kiosk Technologies for your commitment to helping businesses succeed..

[Audio] Two cases of cardless sessions are presented. The first session began at 12:54:20 for the purpose of FORCEDEPOSIT and ended at 12:56:49. The second was initiated at 13:44:10 for POWER-UP/RESET, ending at 15:26:39. An appendix is included with additional details of the case to help decrease Mean Time To Identify (MTTI) and Mean Time to Repair (MTTR)..

12:54:20 Initiating Cardless Session . . . JAM . . . On Upper Transport Or Entry / RETURNMONEY_DEVICE_FAILURE_JAM . . . 12:56:49 SESSION ENDED.

[Audio] Using our ELK+Kibana Elasticsearch solution, mean time to identify and mean time to repair can be reduced. As an example, let us take a look at two customer cases. In the first case, at 12:54:20 a CashDispenserXFSError and a CashDispenseFailure was initiated and the session was ended shortly after at 12:56:49. In the second case, a CardAccepted out of service transaction was initiated at 03:37:30 and the session was ended at 15:26:39. In both cases, the transaction should not have included an 'M-Status: 04'. Our ELK+Kibana Elasticsearch solution offers alerts for relevant users when an issue occurs and scheduled reports to be shared which could prevent such errors..

[Audio] This slide shows the start and end of two cardless sessions represented by timestamps, along with the devices, fitness levels, and M-Status values for each session. It is important to note that for the fitness value, 00, 04, 004, 40000, 4 or any combination of them should not be used, and also that the M-Status should not be 00 or 04. These considerations must be taken into account when using ELK + Kibana Elasticsearch for data storage, Logstash and Filebeat for data processing, Kibana for visualization, setting up alerts for the relevant users when there is an issue, and scheduled reports for sharing. This will help in reducing the Mean Time To Identify and Mean Time To Repair..

[Audio] Cardless Session is initiated with a BNA Fitness of 03 and an M-Status of 02. It is important that neither of these values are 00, 04, 004, 40000, or 4 in any of their positions, and that M-Status is not 00. If CardAccepted is accepted in the same transaction, the session should end with "YOUR TRANSACTION IS SUCCESSFUL" or "TIMEOUT FROM TIMEOUT STATE"; otherwise ChqDAS-Application RefusedChequeReturned will be triggered and the session will end. Implementing this solution can help companies reduce Mean Time To Identify (MTTI) and Mean Time to Repair (MTTR)..

[Audio] Reviewing slide 18 of 28, an example of a cardless session is shown with the device used being a Card Data Management (CDM) device and the Fitness parameter set to 000000. Additionally, a second example is displayed depicting a device set to IDC and also with a Fitness value of 000000. It is important to keep in mind that the Fitness should not consist of any value with 4 in any position, for example: 00, 04, 004, 40000, or 4. The session was initiated and ended at 18:24 and 18:26 respectively..

[Audio] A cardless session on device BNA with fitness status of 000000 was initiated at 18:24:20 and ended at 18:26:49. Case 13 and 14 on device CPM with fitness status of 000000 began at 02:47:30 and ended at 02:50:49..

[Audio] When analyzing Case 15 and Case 16, it is observed that there is no "MONEY_RETAINED_SUCCESSFULLY" in same transaction. This could potentially cause an increase in Mean Time To Identify (MTTI) and Mean Time to Repair (MTTR). Our solution utilizing ELK + Kibana Elasticsearch to store data, Logstash and Filebeat for data processing, Kibana for visualization, alerts for relevant users when an issue occurs, and scheduled reports to be shared; can be used to identify such problems quickly and accurately, reducing Mean Time To Identify (MTTI) and Mean Time to Repair (MTTR)..

12:54:20 Initiating Cardless Session . . . REJECTED_MONEY_NOT_TAKEN_RETRA . . . . . 12:56:49 SESSION ENDED.

[Audio] At 12:54:20 a cardless session was initiated, but was rejected due to money not being taken. The decline code is 58, and the session ended at 12:56:49. Case 19 had a card accepted at 03:37:30, and the decline code is 91, with the session ending at 15:26:39. This shows that careful monitoring of data and decisions is essential when utilizing cardless transactions..

12:54:20 Initiating Cardless Session . . . CashDispenserXFSError . . . CashTaken . . 12:56:49 SESSION ENDED.

[Audio] Today we'll be discussing how our product tutorials can help reduce Mean Time To Identify and Mean Time To Repair. In this slide we see the specific sequence of events and commands related to a cardless session - from the initiating cardless session all the way to the session ending with Case 23 and Case 24 displayed. Here we can see that if a "MONEY_RETAINED_SUCCESSFULLY" or "CashTaken" are not observed in the same transaction, the session will end at a timeout state. We can also see the DEVICE and M-Status noted at 02:47:30, and the session ends at 02:50:49. By using our product tutorials, we can help alert users to potential problems in a timely manner, helping to reduce Mean Time To Identify and Mean Time To Repair..

[Audio] This case allows users to conduct a random search with any given word. This helps stakeholders search and extract reports on any issue, incident, or decline codes and can help them quickly identify and address any issues that may arise. While the details and conditions of this case are still being discussed and worked on, it promises to be an invaluable tool in our arsenal of incident management resources. Thank you..