AWS Public Sector Blog
Three ways government agencies can modernize with IDP powered by AWS
As talent shortages grow, job vacancies are commonplace across governmental agencies. Add the challenge of retaining existing talent and government agencies face a perfect storm. Due to a technology called intelligent document processing (IDP), agencies can accomplish their missions faster—even as vacancies mount.
IDP automatically captures data from documents and data sources, then quickly analyzes and organizes this information for further processing. IDP addresses both talent shortage and job retention by automating mundane tasks and freeing employees for more valuable and rewarding aspects of their jobs. Additional benefits include increased data accuracy, faster time to market, an enhanced user experience, and cost savings of up to 60 percent. Leveraged correctly, IDP can make public sector agencies more efficient and, crucially, more satisfying places to work.
What does IDP look like in practice for government agencies? In this blog post, we explore three different ways Amazon Web Services (AWS) Partner Accenture leveraged IDP—built on AWS technology—to streamline and modernize the work of public organizations.
Use case 1: Making a labor-intensive claims process 50 times faster
At one agency in the European Union that administers benefits for disabled citizens, staff time was increasingly diverted to mundane tasks such as data gathering and entry. First, one agent scanned the relevant data. Next, a second agent read that data to identify which claim they were looking to process. A third agent then extracted the information manually and entered it into the backend system, before sending out a manual notification to the citizen. Only then could a case worker administer the claim. This tedious and labor-intensive process became worse when a claimant failed to complete the entire form, didn’t understand what was being requested, or communicated via a different channel, such as a letter or email.
Accenture knew the agency’s entire workflow could be transformed with IDP. To eliminate the need for a worker to manually read each claim, Accenture implemented computer vision and optical-character recognition with Amazon Textract and Amazon Rekognition. These technologies quickly identify each claim and extract the relevant information. Next, this data, which is stored in an AWS Data Lake, is automatically processed by the backend system via bots that automatically scale based on workload trends, therefore making the solution extremely cost effective.
Due to Accenture’s IDP solution, up to 80 percent of new claims coming into the agency are handled end-to-end without human intervention—at a speed 50 times faster than before. Not only is staff time freed up, but employees now have the resources to tackle the other 20 percent of claims through a more personalized intervention, such as a phone call. And since this infrastructure is flexible, it can scale up and down according to need, saving money during times of lower usage.
Use case 2: Creating a self-learning model to read and store a variety of handwritten documents
Arolsen Archives is a non-profit organization founded to document the lives of victims of Nazi persecution. With 110 million documents and several million other artifacts, it is the largest archive in the world on this topic. Four years ago, Arolsen undertook a digitalization process so people could perform their own searches online. If they relied on volunteers to transcribe the documents, the organization estimated that it would take more than 20 years to complete this mammoth task.
Accenture was called in to help speed up the process. In addition to creating an IDP platform to process the documents, the team created a series of machine learning (ML) models, built on AWS, that could decipher the different types of handwritten documents included among the identity cards, journals, transfer lists, and prisoner schedules. Since going live with the project, the IDP platform has indexed some 18,000 documents and 160,000 names—a 40 percent increase in productivity over manual transcription. And because they are self-learning, the models will improve in accuracy by 10 percent every two-to-three months. Instead of two decades, the Arolsen project is now expected to be completed within two years.
Use case #3: Helping a government benefits agency scale quickly to meet demand
With more than 20 million claims a year, £212 billion in payments, and a raft of services from child maintenance to disability allowance, the Department for Work and Pensions (DWP) is the largest governmental agency in the UK. Five years ago, DWP set out to build what is known as a “digital front door” to accept claims online. What they were not prepared for was the overwhelming response; the increase in demand was so high that they estimated they would need five to 10 times the amount of resources just to keep up.
Accenture knew they could help. Over a five-year period, the Accenture team built 70 end-to-end IDP workflows for DWP, using AWS technology. To date, these workflows have processed 28 million items and saved a whopping three million operational hours, or £60 million. As an additional benefit, its new fraud-detection capacities have delivered tens of millions of pounds in additional savings.
When the pandemic arrived and the UK announced a countrywide lockdown, DWP received one million claims in the first 24 hours. Thankfully, because Accenture had built the workflows in the AWS Cloud, the system had no problem scaling up to meet these additional claims. And when the government added a series of new benefits to service the most vulnerable during the crisis, Accenture quickly built an end-to-end automation service to channel these new claims through the platform.
IDP for today’s needs
Many public agencies today are behind on IDP. According to Campbell Abbey, global public sector lead for Accenture’s AWS Business Group, previous generations of automatic document processing failed to deliver, creating a reluctance to move forward. But today’s IDP, built in the flexible and scalable AWS Cloud, has the potential to revolutionize agency workflows.
“We have to move beyond the mindset of automation being about siloed individual tasks performed by bots, and put a rigor of orchestration intelligence around them,” he explains. “Build holistic, end-to-end automation projects that solve for the workforce and not just the individual. And have a centrally managed solution able to run intelligent document processes with modular designs so that we can build and scale them both faster and cheaper than previous generations.”
That is precisely how AWS services such as Amazon Textract and Amazon Rekognition apply ML to IDP solutions. The new generation of IDP is here.
Learn more about implementing IDP in your agency by exploring the AWS Solutions Library. See how more than 7,500 government agencies around the world use AWS at the AWS for Government hub.
Accenture and AWS have worked together for more than a decade to help organizations realize value from their applications and data. The collaboration between the two companies, the Accenture AWS Business Group (AABG), enables enterprises to accelerate their pace of digital innovation and realize incremental business value from cloud adoption and transformation.
Read related stories on the AWS Public Sector Blog: