Case study on SIB

Sketches how PiXL.AI helped a renowned bank of South India to overcome its issue. Anybody who has been to banks would be knowing about the amount of formalities one needs to undergo. Similar was the case with one popular bank of South India. Each time a person opens an account with this bank, he submits a copy of his documents and identity proofs like aadhaar card, pan card, driving license for the documentation process. Other than opening the accounts, banks perform numerous other essential services for its customers which further requires different kinds of documents.

Case study on SIB
CASE STUDY ON SIB

The bank maintains and stores scanned copies of such documents which are saved as images.It needs to classify, extract and segregate the documents for its further usage and record keeping purposes. These works were carried out by the bank’s employees. Firstly, the bank’s employees were required to look into each image, classify them into various groups for instance all the adhaar card would be put into a different category than a pan card. Once this has been done, the next step would be to extract all the important data from these categories. The third step would be to segregate the extracted information according to the need in order to ease down the process.

Such big banks usually deal with the large amount of data collection on a daily basis which would accumulate to even bigger proportions. The data with the bank was bulky and amounted to 16 millions. Now the bank was in the dilemma to do the entire classification and extraction of documents using human resources available with the bank or outsourcing this service to any vendor who would be using their technology in order to fulfill the necessity.

In the first case where employees of the banks work manually to meet the requirement it would take six months or even more than that before they can work out on the three processes and then actually extract the information. Even after putting so much efforts the bank would be running on the risk of having potential errors. Thus, the accuracy level would be very low and risky since humans tend to make errors. The second drawback of this choice would be the expenditure. The cost which the bank would incur to do this grind would be tremendous. Due to these reasons the bank did not choose this option rather they preferred going forward with the outsourcing of work.

So, the bank chose the second case and starting looking for various vendor’s services. The bank tried out various vendor’s services but each time it was dissatisfied with the work done as the vendor’s were lacking in one of the major key points. It was the accuracy. The banks could not put their credibility at risk. Those vendors could provide only 80% of the accuracy for the entire process. Since even a small error can turn out to be detrimental to the organisation due to the significance and sensitivity attached to such documents, the organisation can’t afford to have such less accuracy in work in order to carry out its operations seamlessly.

Lastly the bank came across our product which was exactly what they were in search of all this time. Our product PiXL KYC is an AI based image classification and data extraction system. It provides top notch technology in image processing. PiXL KYC started the job with firstly classifying the data into various categories. It would do so by reading each image pixel by pixel, doing so it would be able to identify the document type. It would also make a quality check for the documents. It there is any discrepancy with the quality required it would be reported to the bank. PiXL KYC is able to automate customer identification and verification process and link documents and the respective user account. This step would ensure the banks of the safety as it would be able to detect a potential fraud or mismatch in the documents. Next, it would extract all the useful information from the pool of data available in the form of images and documents. PiXL KYC has superiority over all the other’s vendor services able to provide 98.2% of the accuracy, which the highest when compared to any other vendor’s product. The rest 1.8% accuracy loss is because of the corrupted documents.

The entire process on 40 millions of documents was completed in less than a month which would have otherwise taken 6 months or above.

We are proud to announce that our product PixL KYC could play such a crucially vital part in automating digital documents classification and extraction or segregation. The automated process ensures the elimination of the errors which a human labour could have caused, this saves the bank from the unnecessary fuss and provides a smooth delivery of the functions with a surge in satisfaction level on the part of employees as well as the service takers of the bank. This helps in increasing the efficiency. We provided them with the best solutions. Our services helped the bank in not only saving time and providing the best quality but also ensured a great cost cutting for them. Our service is highly cost-effective and can be afforded in a most reasonable price. therefore , even the most price sensitive businesses need to worry. Imagine the amount of time, manpower and money would be needed to extract information from 16 million data manually. In a place such as bank where each moment is valuable saving 5 months in the work process was of immense significance. Our services are available to every customer who deal with documents in the form of images and find it hard and time-consuming task to extract data from it. We will do it for you.