March 26, 2021
Stop manually entering data from ACORD forms. Humans are slow, error-prone and expensive. It’s time to automate your document-heavy Insurance business processes using Optical Character Recognition (OCR) and Automated Machine Learning (AutoML) technology from Impira. Your customers will thank you.
Insurance companies remain heavily reliant on paper-based, manual processes, even as consumer expectations shift around them. In the post-COVID-19 world, Insurance companies now realize they must prioritize AI-powered automation if they wish to compete on customer satisfaction.
Many of the time-consuming administrative processes in Insurance involve handling ACORD forms to capture customer input. ACORD forms are governed by a nonprofit organization called the Association for Cooperative Operations Research and Development. ACORD forms and certificates are used by 90% of property and casualty insurance carriers in the U.S. Examples include:
For a full list of available ACORD forms, see here.
Even in 2021, Insurance companies are receiving these ACORD documents via email, fax, and snail mail. Personnel must devote much of their time to manually typing this information into some computer software. To alleviate this pain, many companies have attempted to use traditional OCR technology, only to find that this legacy technology (with no AI intelligence) creates so many errors that they revert back to doing things by hand.
Impira AutoML technology combines OCR with artificial intelligence and machine learning to fully realize the dream of document automation. With Impira, you can easily turn your ACORD forms into business ready data in seconds. Unlike OCR of the past, Impira AutoML learns from mistakes, continuously improving performance and data quality.
Impira AutoML can be trained to learn any ACORD form or any other kind of document used by your business in just a few clicks. How? Impira AutoML does not rely on pre-trained models. We provide an easy-to-use interface where users can upload a few document samples and highlight the text they’d like to extract. Our system will learn from your highlighting interactions and immediately output the results. Let’s give it a try using the ACORD 25 Certificate of Liability Insurance form. If you have some examples handy, feel free to use your own. Otherwise, you can download samples from here.
Impira dramatically reduces the amount of time and effort required to process documents. With Impira, teams can get more work done in a shorter amount of time. This results in cost savings to the business. Additionally, faster processing times are proven to boost customer satisfaction. This translates into higher revenue and reduced customer churn.
Impira AutoML approach to document processing means that there are many more use cases that can be solved within your insurance organization than the one described above — from electronic payment authorization, to contract change request, lab reports, withdrawal forms, and more. Give it a try on your own by walking through the steps outlined below.
If you haven’t already done so, head on over to impira.com/signup to create your free account. There’s no credit card or personal information needed to get started beyond an email address. Our free plan includes 200 file units (pages), so there’s no cost to follow along with this tutorial.
Once you’ve created your account, start by creating a new “collection”. A collection is the main feature within Impira for organizing and grouping files together. Collections are a lot like folders on your computer. Click on the ‘plus’ symbol next to the word “collections” in the left-hand sidebar.
In the dialogue box that appears, give your collection a meaningful name of your choice, such as “ACORD 25”.
Click on the name of your newly created collection.
Then, drag and drop your sample ACORD 25 forms onto the browser, over the text where it says “Upload some files”. Set aside one sample file to be used at the end of this tutorial.
Your files will immediately begin processing. When complete, you will see that each file has its own row in the table, as shown below.
Double-click on the first file to open the Impira mark-up interface. Let’s start by extracting the “Contact Name” field. Simply click on the contact’s first name and drag your mouse to highlight their full name. On the right-hand side bar, label this field as “Contact Name”. If you close the mark-up interface by clicking the “X” in the top-right corner, you’ll see that we’ve added a new column to our table for this attribute.
Notice the spinner in the column header. This indicates that the Impira AutoML is updating the model under the hood. It then immediately applies this learning to all of the documents in your collection. In just a few seconds, we can see the contact name for every ACORD 25 form we uploaded.
Notice that some of the cells may still be null and have a red low-confidence flag next to the left-side of the cell. This is a quick visual indication of the underlying numerical confidence score for the prediction. We need to provide additional training to the model to boost the confidence levels. Double-click on a low-confidence cell to reopen the mark-up interface.
Impira has drawn a box around the predicted text on the page. If the value is correct, simply click the “Confirm value” button on the right-hand sidebar.
If the value is incorrect, remove the predicted value by clicking the “X” on the field name, and redraw a box around the correct value.
Each time you make a validation or correction, that information is fed back into the AutoML model for immediate re-training. Validate a few more predictions, and you will start to see the confidence flags change from red to green, indicating high-confidence.
At this point, you will likely want to add some more fields to collect the remaining information on the ACORD form. Go ahead and repeat the above steps to add fields such as Phone, Email, Producer, Insured, NAIC #, etc. Here are a few extra tips:
As you add these additional fields, the number of columns in your table will grow.
Once you are satisfied with your extraction results (i.e. all the cells are displaying the green, high-confidence flag), you might choose to download your data as a CSV file. To do that, expand the drop-down menu labeled "Download" in the top right hand corner, and select "All files records (CSV).
Alternatively, you may want to programmatically access your extracted data. Impira provides a robust set of RESTful APIs for both reading and writing data. Now that your AutoML model has been trained, any new ACORD 25 forms that are sent to the Impira platform will be automatically categorized and extracted.
Remember the one sample file we set aside in Step 2? Let’s imagine that you receive this file for intake at a future point in time. As part of your workflow, this file can be ingested into Impira through various means (e.g. Dropbox sync, S3 sync, etc.). For now, (to simulate this) drag and drop this remaining file onto the Collection. Going forward, new ACORD 25 forms sent to Impira will be appended to this Collection as a new row. The Impira AutoML will then automatically be applied to this new document, and the information made available to downstream systems.