Loan Application Process Automation with AWS ML and UiPath

Many organizations start using Robotic Process Automation (RPA) to automate workflow and/or back-office processes that are labor-intensive.

In this post, we show how a financial institution can use a combination of UiPath and AWS ML services to automate a loan application process.

UiPath is a Robotic Process Automation platform for end-to-end high-scale automation. UiPath software offers solutions for enterprises to automate repetitive office tasks for rapid business transformation.

Robotic process automation (RPA) is a software technology that makes it easy to build, deploy, and manage software robots that emulate humans’ actions interacting with digital systems and software. Robotic process automation streamlines workflows, which makes organizations more profitable, flexible, and responsive. It also increases employee satisfaction, engagement, and productivity by removing mundane tasks from their workdays.

Amazon Textract is a machine learning (ML) service that automatically detects and extracts text and data from scanned documents. It goes beyond simple optical character recognition (OCR) to also identify the contents of fields in forms and information stored in tables. Amazon Textract uses Optical Character Recognition (OCR) technology to automatically detect printed text, handwriting, and numbers in a scan or rendering of a document, such as a legal document or a scan of a book.

First, we show how to automate the applicant’s personal and financial information validation using Amazon Textract to perform OCR and extract valuable information which are store in AWS DynamoDB. Second, we perform an automated form and data validation using UiPath Robotic Process Automation bot.

Process Overview

The following architecture demonstrate step by step workflow of the process and the interaction between Textract and UiPath for the loan application process automation.

  1. Client accesses the financial institution loan application website, complete the form by providing personal information and upload required documents such diver license, social security card, W2 forms and 1040 tax return and submit the form.
  2. System stores the form data in database and upload documents into AWS S3 bucket after validation and scanning for vulnerabilities.
  3. UiPath Orchestrator starts the process workflow and download documents from S3 bucket and invokes Amazon Textract service to analyze documents and extra key value pairs data which are stored in database.
  4. Documents are scanned for vulnerabilities and uploaded to AWS S3 bucket.
  5. A business process is invoked to run business rules after retrieving data from database. That includes form data and document information matching, credit validation and internal financial rules for decision making
  6. Approval or rejection decision is generated from previous step and a notification email is sent out.
  7. If the loan is approved, an email is sent to the applicant with next steps.
  8. If the loan is rejected, the email is sent to the loan officer for review and follow-ups.
  9. If the loan is approved, loan documents are generated from the document management system and users and financial information are recorded in the financial institution databases.

Krai Academy Solutions Machine Learning and Deep Learning Initiatives

Much of health data today is in free-form medical text like doctors’ notes, clinical trial reports, and patient health records. Manually extracting the data is a time-consuming process, while automated rule-based attempts to extract the data do not capture the full story as they fail to take context into account. As a result, the data remains unusable in large-scale analytics needed to advance the healthcare and life sciences industry and improve patient outcomes and create efficiencies.

At Krai Solutions, we have built a Machine Learning Cloud based platform with is deployed either a SaaS or subscription web-based model to allow with a simple API call to quickly and accurately extract information such as medical conditions, medications, dosages, tests, treatments and procedures, and protected health information while retaining the context of the information. The platform is built on top of AWS Amazon Comprehend and can identify the relationships among the extracted information to help you build applications for use cases like population health analytics, clinical trial management, pharmacovigilance, and summarization. You can also link the extracted information to medical ontologies such as ICD10-CM or RxNorm to help you build applications for use cases like revenue cycle management (medical coding), claim validation and processing, and electronic health record creation.

Benefit of the platform:

Understand and identifies complex medical information quickly and more accurately. For example you can extract "methicillin-resistant Staphylococcus aureus" and provide context, such as whether a patient has tested positive or negative to a medical procedure. Provide several capabilities to help healthcare providers stay compliant and protect patient data by implement data privacy and security solutions by extracting and then identifying relevant patient identifiers as described in HIPAA’s Safe Harbor method of de-identification.

Use cases:

1. Perform medical cohort analysis: In oncology, it is critical that the right selection criteria are quickly discovered to recruit patients for clinical trials. Our platform understands and identifies complex medical information found in unstructured text to help make indexing and searching easier. You can use these insights to identify recruit patients to the appropriate clinical trial in a fraction of the time and cost from manual selection processes.

2. Support clinical decisions: The platform extracts medical information from patient data stored in Amazon S3 and returns structured results that you can integrate into a healthcare dashboard a care support team can access. The dashboard may contain built in early warning system to help identify individuals at risk of multiple sclerosis by extracting diagnosis, sign, and symptoms from more than 100,000 clinical notes Improve medical coding in revenue cycle management: For a hospital, the process of finding the right diagnosis in the patient notes that should be mapped to the correct code in the International Classification of Diseases (ICD) can be time-consuming and tedious. It is particularly challenging to extract diagnoses that can be represented in different ways. For example, “atrial fibrillation” is sometimes written as “AF.” With a call to our platform API, you can accurately identify abbreviations, misspellings, and typos in medical text. This reduces the time a medical coder must spend analyzing unstructured notes, decreases the time burden on clinical staff, and improves efficiency.

To learn more, please click here Krai Solutions AI

Code Creator LLC

As Amazon Webservices expert partner: Code Creator LLC is a software provider for Amazon Web Services and AWS Marketplace. They create cloud applications, compiled, and configured in an instant, ready to run state for users and developers. Code Creator builds large portions of the software toolbox and makes it available to anyone and everyone who might find it useful, helping firms of all sizes reach a new level of productivity, whether they've invested heavily in their IT department or not. In this project the Code Creator was looking for an OCR solution that is deployed in AWS Cloud as SaaS solution using AWS Textreact. Amazon Textract is a service that automatically extracts text and data from scanned documents. We have implemented a solution using AWS Serverless architecture composed of Cognito, S3, EC2, Lambda, Textract and Elasticsearch as described below:

Overall, our client was impressed with the design approach and the deliveries I produced and open to the idea of future collaboration and support from me. After the client received our first design proposal here is the initial response and I quote “I think you are the only one of all the submitted proposals who gets it.” It was a pleasure to work with this client and we are looking forward to continuing to deliver best in class solutions to clients on AWS IQ.

National Institute of Health, NIH

As a Subcontractor for Metro Star Systems:

  • Architect e-GOS Amazon Cloud Infrastructure and Systems Setup and Configuration using Cloud Formation Templates. System is Fed Ramp compatible.
  • Perform database administration duties for the e-GOS application, including but not limited to capacity planning, installation, configuration, database design, data migration, performance monitoring, security, troubleshooting, and backup and data recovery tasks.
  • Perform basic system administration duties, including but not limited to: installing, supporting, and maintaining e-GOS servers and planning for and responding to service outages and other problems.
  • Work with subject matter experts to analyze the e-GOS code base in conjunction with database, infrastructure, and development environments to assess performance, stability and security. This includes:
  • Work with subject matter experts to implement database and system improvements.

Metrics: Platform is composed of 5 environments (DEVELOMENT, QA, UAT, DEMO, PRODUCTION) AND 4 tiers: front end using Angular JS, back end using Spring-Hibernate, Data layer using MS SQL and integration tier with Apache SOLR. Production and UAT environments are composed of a cluster or servers which are site behind AWS Elastic Load Balancers (ELB) and auto managed with auto scaling groups. The database is composed of about 240 tables, 470 triggers and about 5,000,000 data rows.

Schneider Electric, eCommerce re-platform

As Senior consulting of McFadyen Solutions, provided services to design and develop a worldwide (multi language and multi currencies) Ecommerce application enabling effective customer experience and streamlining business processes. The application is built on top of Oracle Commerce Platform with some customization thereby increasing Revenue, Conversion rates, Average order value, and Traffic. The system integrates with a content management system to enable eCommerce application site content to be managed by Tridion, SDL Content Management System (CMS) and streamline business processes thereby building a foundation for more Schneider Business Units to integrate and share content seamlessly. The architecture is a multi-layers, multi-tiers system offering the benefits of performance improvement, scalability, extensibility, maintainability, and manageability.

Metrics: Schneider Electric, eCommerce platform is a multi-site multi-language eCommerce solution covering 173 languages. Each site has its own catalog that is derived from a master catalog where each language maintains its copy contents of the catalog data in a separated repository providing the content localization which includes translation and currency conversion. Because of country specific tax policies, the platform needs to integrate with multiple tax services and payment gateways.

Teleflora, eCommerce re-platform

As Senior consulting of McFadyen Solutions, designed and implemented a federated multi site eCommerce application using Oracle Commerce Platform. The design and implementation used industry best practices and standards and provided effective customer experience and streamlining of business processes thereby increasing Revenue, Conversion rates, Average order value, and Traffic.

The architecture is a multi-layers, multi-tiers system offering the benefits of performance improvement, scalability, extensibility, maintainability, and manageability.

Metrics: The architecture is a multi-layer, multi-tiers system offering the benefits of performance improvement, scalability, extensibility, maintainability, and manageability. The application platform is a multi-site eCommerce solution with about 15,000 sites. The master catalog has about 4,000 SKU. Each site has its own catalog that is derived from a master catalog where each site can override any property of category, sub-category, product, or SKU. This can bring the number of SKU to be close to 10-60 million. Teleflora e-commerce solution is a billion dollars business with about a million users.

Direct Brands, Inc eCommerce Design and Implementation

As an employee of Direct Brands, Inc, designed and implemented an eCommerce solution using ATG platform which integration to Legacy Systems using SOA. The solution offered the benefits of improving user experience and streamlining business processes thereby increasing Revenue, Conversion rates, and Average order value. The architecture is a multi-layer, multi-tiers system offering the benefits of performance improvement, scalability, extensibility, maintainability, and manageability.

Metrics: Direct Brands eCommerce platform is a subscription based multi-site application with about 22 books and DVD clubs. There about 2 Million subscribers with more than a million SKUs.

About Us

Krai Solutions helps organizations transform and enhance their business processes using IT best practices such as Agile, DevSecOps and innovative approaches to solve challenging problems..

Get in Touch

How to Find Us