Intelligent automation and lending glossary

Auto lending is the process of setting up a loan for an individual in order to finance the purchase of a vehicle. The terms of these types of loans may vary from borrower to borrower.

Reviewing human-written documents with computer programs instead of people.

A part of a company that performs non-client facing functions and personnel like adminstration and support.

Valuable information and insights from a complex banking data set that consists of both structured and unstructured data.

When a criminal steals a company’s identity by creating, using or attempting to use the identifying information of a business, without authority, to buy goods, services and/or obtain tax benefits.

BPM is the ability to utilize a variety of methods to continuously analyze and improve upon current business processes. Organizations often take advantage of automation and technology to aid these efforts and become more efficient and effective.

Valuable information and insights from a complex data set that consists of both structured and unstructured data.

The examination of different cash inflows and outflows of a company or entity to get a detailed picture of its financial health. Ocrolus offers 100+ small business and consumer analytics that illuminate a borrower’s cash flow dynamics, including revenue, expenses, transaction categories, financial ratios, and trends.

The practice of adding metadata to documents that makes them easier to search for.

Employment verification process of confirming a borrower’s work history with documents such as W2s, paystubs, and Work Number forms.

The process of identifying multiple data sources that refer to the same entity and then linking the records together. Entity resolution can be especially useful in the case of reconciling customer information when their data is entered multiple times in unique sources by various people.

This is the practice of creating a contractual agreement where the lessor, who owns the equipment, allows the lessee to use the equipment for a specific amount of time and for periodic payments. The equipment may refer to vehicles, factory machines, or any other equipment.

The deliberate misrepresentation of account balances in a financial statement to help an individual or organization deceive the person or entity reviewing the financial statement, such as a lender. This misrepresentation often helps individuals gain access to more credit than they normally would.

Converting the contents of forms into machine-readable data. Depending on how the form is stored, data extraction can be done through OCR or through the document’s embedded text data.

These are human-made unplanned mistakes that reduce quality, safety and/or security.

An extension of machine learning in which human experts validate the analysis performed by machines. This combines the best of machine learning with the review of humans.

Identity fraud is when an individual pretends to be another person using stolen or forged documents in their name.

The verification of how much a borrower earns through the assessment of documents like bank statements, paystubs, and tax forms. Ocrolus can calculate and verify income for consumers with diverse employment profiles.

A type of document automation where a program is able to capture, extract, and process data from a variety of document formats. It utilizes AI technologies such as natural language processing (NLP), Computer Vision, deep learning and machine learning (ML) to classify, categorize, and extract relevant information, and then analyze the data.

This is a standard in the financial industry that ensures financial professionals have detailed information about their clients’ risk tolerance, investment knowledge, and financial position. KYC protects all parties in the financial process by making sure both sides are aware of what is a reasonable financial burden and risk. KYC compliance typically involves requirements and policies such as risk management, customer acceptance policies, and transaction monitoring.

A system that automates and manages the loan process from beginning to end. This includes everything from the application, through to underwriting, approval, documentation, pricing, funding, and administration. While the individual steps may differ with every lender, the overall process is very similar and helps to maintain a lending relationship.

The practice of getting approved for multiple loans or lines of credit at the same time or within a very short period of time. Loan stacking generally happens through online lending and can be done by either individuals or businesses. While this practice is not illegal, billions of dollars are lost by financial institutions each year because many loan stackers commit application fraud and intentionally default.

The practice of building a computer model to solve a task by letting it discover patterns in data, rather than by using traditional programming methods and setting rules in advance. Furthermore, one of the distinct advantages with utilizing machine learning is that these models can be set up so that with each performance of the task, the computer is able to learn how to perform the task better.

The technology stack of a financial service company that helps streamline processes, increase accuracy and reduce turnaround times. Currently, the fintech field is broken up into four main categories, 1) B2B for banks, 2) the banks clients, 3) B2C for small businesses, and 4) consumers.

This is the process of replacing manual data entry with automation in the mortgage process from start to finish. Mortgage automation significantly enhances the lending process by enabling financial institutions to route forms and data automatically, reducing the complexity of processing applications, decreasing potential errors, and ensuring customers an efficient and pleasant lending experiences.

The process of ensuring that the mortgage process proceeds according to regulations. This includes preparing the right documents, getting signatures from the correct individuals, providing the proper disclosures at the correct times, and maintaining comprehensive documentation.

The ability of machines to convert images of typed text into machine-readable data.

RPA is the process of designing a machine (including software) to do human tasks using real or simulated human tools.

This index is a measure of the health of small businesses in the United States. This number is based on several different sets of data including Ocrolus’ own and seeks to capture a real-time snapshot of the health of small businesses by industry and geography.

Synthetic identity fraud is the practice of pretending to be a non-existent person using partially or entirely false information. This is in contrast with conventional identity fraud, in which the identity of a real person is assumed.

Transactional data is the data gathered as a result of an individual or businesses transactions. Recent transactions tend to provide more useful information than older ones.

The underwriting process is initiated when an individual or institution is interested in some sort of financial risk, namely taking on a mortgage or insurance . In order to confirm eligibility, the lender verifies the individuals income, assets, debt, credit and property.