How AI is Used to Audit Bank & Credit Card Statements
Artificial intelligence has been immensely advantageous to the financial sector. The lengthy tedious work which took hours for humans to perform was reduced to seconds by computer software. The use of papers, pens, and abacuses for conventional accounting systems have been transformed into the computerized system of accounting and auditing. Innovation in this field is so powerful that it has elevated the financial auditing process to a higher plane.
Ocrolus is a financial service provider with two powerful AI products: “PerfectAudit” and “MedicaidGenius.” These products eliminate the need to audit bank and credit card statements manually.
The company was founded in 2014 and is headquartered on Wall Street in New York. After testing and polishing, Ocrolus launched officially in January 2016. Its team consists of four executive members–Sam Bobley, Victoria Meakin, Vikas Dua, and Zoheb Sait. There are an additional 15 employees working on the technology and operations side of the business.
Victoria Meakin serves as president of the company. She was also the co-founder and president of PhoneCharge, an electronic payment processing company that sold to CheckFree for around $100 million. CEO Sam Bobley is a young technology entrepreneur who graduated from the University of South Carolina-Columbia. COO Vikas Dua was previously associated with on-demand Series-C startup Handy.
What Information PerfectAudit Analyzes
PerfectAudit has been described as a fintech platform with artificial intelligence built into an audit methodology in order to perform complex processes. It collects bank and credit card statements for previous years in any format including PDF, e-statements, fax, scanned originals, and photocopy. The platform promptly extracts and filters all the data by accounts, date, description, and dollar amounts for an easy approach. If any data is not automatically ratified, then it is snipped into small images for the precise value. Taking it a step ahead for advanced accuracy, data is crowdsourced to pre-qualified workers for manual data validation simultaneously to ensure that any personal information is not shared. The beginning and end balances of account statements are reconciled with the sum of transactions through an algorithmic reconciliation process. The data is exported in the encrypted form and stored with Amazon Web Services. The combination of Optical Character Recognition (OCR) transaction detection, crowd-sourced data validation, and algorithmic reconciliation has proved to be a winner with more than 99% accuracy.
The platform does industry-specific analysis in computing series of results such as average daily balance, list of the day’s negative balance, list of interbank transactions, list of deposits from outside sources, average deposit by month, max deposit by month, minimum deposit by month, average transaction size, list of NSF transactions, list of alternative lender transactions, list of transactions ending in three zeros or more (0.00), and check to ensure previous month balance is the same as beginning balance of next month. This helps the client to conclude whether the decision to lend should be made or not.
The digitalized financial outcome is made conveniently accessible to clients through a user-friendly interface where all the desired and relevant information can be searched at a click of the mouse. It requires no installations or upgrades as everything is uploaded to the cloud, which is much more reliable and secure than downloaded files.
How Ocrolus Products Facilitate Good Lender Decisioning
PerfectAudit helps control the risk of bad debts and frauds by meticulous decision-making process and with quick automated data validation. Lenders are able to scale loan volume with reduced human intervention. The solution furnished consists of a unique standardized process for every lender. It provides a single platform with processes starting from data collection, filtration, and validation till financial determinations are made. This cycle is incessantly performed, which requires consistent accuracy every time it is done. AI excels in performing these repetitive tasks.
This platform ensures quality, effectiveness, efficiency, and insights in providing complex data reports. It replaces prolonged tiresome manual work with incredibly fast, inexpensive, and reliable solutions. It gives a very high rate of accuracy even when the records are suspicious and fuzzy. Auditors generally use a sampling method in the auditing process while PerfectAudit analyzes 100% of documents and leverages ready-to-use material, which gives an opportunity to the auditors to have in-depth insight into advanced queries to reach improved judgments.
The company’s proprietary technology also allows for regulatory compliance. It helps document the process of credit decision making, and every decision can thus be built backward onto the central algorithm.
The company’ s second product, MedicaidGenius, is targeted towards the Medicaid processors facing the same issue.
Ocrolus is Positioning Itself to Lead the Fintech Niche
The company initially focused on working with family lawyers and now works for auditors, accounting firms, and government departments. dealing with overwhelming documentation work. The company is now prioritizing the lending industry, and fintech lenders are on its radar.
Ocrolus has successfully provided services to more than 100 clients. The biggest is Strategic Funding Source, and other major clients include LendStreet Financial, Lendified, Oscher Consulting Inc., the law office of Gold and Gold, Nolan Stewart PC, Mercer Capital, and Largo Police Department. The company offers customized payment models to fulfill clients’ requirements. Most of the lender clients prefer monthly or quarterly bases, but a few prefer payments-on-time and per-page bases.
The cost-effective tool of Ocrolus has cut short the page-by-page examination into a streamlined, hyper-accurate process. The technology combines the best processes of previous solutions like OCR and combining them with its patent pending IP-in-transaction detection. It also leverages crowdsourcing and revalidation to ensure further accuracy. The automation of data validation saves time and money and eliminates the possibilities of human error.