What does automated underwriting mean? It generally refers to automating tasks, using artificial intelligence and machine learning algorithms to help make decisions, and taking advantage of other automation opportunities. In practice, automated underwriting can mean different things to different people.
Many processes have an underwriting requirement, including:
- Loan origination
- Insurance policy acquisition
- Bank accounts
- And more
The key to underwriting is to make data-driven decisions about the worthiness of applicants for insurance, bank accounts, loans, etc. based on certain evaluation criteria. Implementing an automated underwriting system or process is a perfect way to augment manual underwriting processes where humans are responsible for assessing:
- Financial information
- Credit history
- Risk profile
- Other applicant-relevant information
Examples relevant to the above would include automated mortgage underwriting, automated loan underwriting and automated insurance underwriting.
Combining the automation of manual tasks with the ability to analyze extensive amounts of data from a wide range of sources helps organizations streamline and optimize the evaluation of applicant eligibility. By automating the process, you can swiftly and effectively handle large application volumes, resulting in faster decision-making and reduced administrative expenses for lenders and insurers.
Benefits of automated underwriting
To help you better understand automated underwriting and what it can do for your business, let’s take a look at some of its key components and features in the following table.
The underwriting process can benefit from automation in several areas. Many of these are quick to implement and some take more research and time. Here are some options to consider:
| Component | Description | Process Orchestration/BPMN Example |
| Efficiency and speed | One of the main reasons to automate any process is to make it more efficient—which often results in overall faster processing. In addition, automating manual tasks often reduces the risk of human error, adding to the efficiency of your process.Consider automating any manual or decision-making tasks first. Taking this approach can significantly reduce the time required for underwriting processing, which will mean faster turnaround times for applicant approvals. | If, for example, you have a human task to calculate the debt-to-income ratio (DTI) or review an insurance or credit score on an application, this is a task that you can easily automate.Calculating DTI and passing this and other information to perform an automated credit score look-up is a common set of operations. The resulting information can be used in a Decision Model and Notation (DMN) table to provide a risk assessment for the applicant. This will both minimize human error and expedite processing. |
| Risk assessment | One of the strengths of automated underwriting is the ability to assist with risk assessment.Many automated underwriting processes use risk assessment to help determine the appropriate insurance premium for an applicant or even whether default may be an issue. An applicant’s credit score may help you decide to issue a higher interest rate, a lowered credit limit, or even no loan or credit card at all. | To automate risk assessment, submit the applicant’s information to an automated system to obtain their credit score or an insurance score.For example, you might create a GET request through a REST Connector to access an external system that will return the score information requested. |
| Rules-based decision-making | Traditional underwriting processes often have several predefined criteria and rules regulated by the industry, as well as your internal business procedures.These rules help you evaluate the eligibility of applicants and their associated risk levels. Eligibility differs by industry but often includes DTI ratios, credit scores, salary and net worth, past payment history, and insurance score. | Many companies have established rules that change periodically when regulations in the industry change or internal policies change.For example, your company may set a minimum credit score of 675 with a DTI of less than 36% as a requirement for a credit card with a $10,000 credit limit.This is a perfect construct to use a DMN decision table. The table can include multiple inputs (credit score and DTI) and output a total risk score for the applicant.By extracting this decision logic from the process, it can be easily updated if the rules or regulations change without the need to alter the underwriting process. |
| Automated notification | You often need to communicate the status of the underwriting process to an applicant, or you may need to request additional information. By automatically sending updates and notifications, you can improve customer service and efficiency. | With Connectors for messaging platforms—such as email, Slack, and SMS—your underwriting process can include automatic customer communications. A customer can automatically receive a status without waiting for a human in your organization to generate the message and send it. |
| Generative artificial intelligence (AI) | Taking advantage of generative AI can speed up the creation and implementation of your underwriting process. You can use it to provide text for automatic notifications.Your process developers can also use generative AI to help build processes more quickly. | Consider taking advantage of AI-generated forms to create a simplified interface allowing for easy input from your clients.You can also use generative AI to create text for automated notifications. For example, use a Connector to OpenAI to pass certain process instance specific information to obtain process appropriate content. Using this generated message, you can provide automatic status updates or information requests. |
| Updating legacy and systems of record | Underwriting processes often require integration with other systems. In these cases, you want to automate the pushing or pulling of information from these systems, avoiding manual entry.This can help you maintain the appropriate system of record, as well as minimize human error while expediting system updates. | Use connectors, webhooks, service tasks, or other BPMN constructs to read from or write to systems of record. Each construct provides a different method for making updates, but each will minimize errors associated with manual updates and improve operational efficiency. |
| Automated escalation | Underwriting processes often require escalation. Maybe you are waiting for the return of information from the applicant or documentation and the waiting interval has passed or the applicant’s situation requires special handling and needs the attention of a supervisor.Automating this escalation will improve both customer service and the efficiency of your underwriting process. | If a wait threshold is reached, the process can automatically route the application to another level or send a reminder to the applicant. This eliminates the need to manually make these adjustments.You can achieve this using timer events or escalation events, depending on the circumstances. |
| Data integration | Most underwriting processes require the consolidation of data from multiple sources, including credit reports, financial statements, employment history, credit scores, and policy declarations.Automated data integration removes time-consuming, error-prone manual procedures, allowing organizations to merge, manage, and derive insights from their data resources with greater efficiency. | When evaluating an applicant, you may need to review a wide range of information found in various materials from various places.To integrate this data, take advantage of things like artificial intelligence and optical character recognition. This might be in the form of calling a service, passing it a document or audio recording, and asking that service to locate key information.By locating or extracting or locating the information required to make a quality decision, you can make your underwriting process more efficient. You’re providing your staff with only the information they need without the time-consuming struggle of locating each piece of information. |
| Machine-learning algorithms | Machine-learning algorithms can analyze historical data and identify patterns to improve decision accuracy and risk assessment over time. These algorithms adapt and learn from new data to continuously refine and optimize your automated underwriting process. | As you process applicants, you can use process data—both historical and real-time—to review patterns and provide insight into future volumes and issues.You can also review historical process data to see if you’re meeting your KPIs. For example, if you’re promising applicants that their loan can be issued within a certain number of days, you can use the process data to determine if you’re meeting this KPI.This data can also be used in combination with other data within your organization to improve your accuracy and optimize processing. For another example, if you find that home values have been decreasing in a specific area over the last several years, you may need to adjust your underwriting procedure to accommodate those lower values. |
| Scalability | By automating your underwriting process, you’re enabling it to handle volume fluctuations in your applications. The ability to scale like this allows you to efficiently manage unexpected spikes in demand without compromising your decision quality or customer service. | Most financial and insurance companies experience large variations in application volume over time. Automating manual tasks and risk assessment enables you to handle increased volume without additional staff or incurring human errors. |
Clearly, you can realize so many benefits from automating your underwriting process, like enhanced decision accuracy, reduced costs, better customer service, and overall efficiency.
If you do decide to start automating your underwriting process, just be sure to follow all regulatory requirements, fair lending practices, and ethical standards. This will ensure that you can reap the benefits of automation while avoiding pitfalls around bias and discrimination.
Automating processes is one of the best ways to stay competitive in today’s marketplace.
Learn more
Learn more about how Camunda can help you with your automated underwriting needs with a free trial. Camunda is a powerful and flexible process orchestration platform that can help you automate your underwriting processes and drive lasting value.
Keep reading
Want to dive deeper into how automation can help with underwriting? Check out these blog posts
Start the discussion at forum.camunda.io