Amazon Web Services Inc. today shared details about how a few of its customers have used Amazon Textract and other artificial intelligence offerings in its public cloud to respond to operational changes during the pandemic.

Amazon Textract is a managed AI service in the AWS cloud that can extract text from documents. It’s a specialized capability, but one that is applicable to a broad range of use cases because virtually every organization deals with documents as part of its operations. This broad applicability is part of why AWS’ rivals have also introduced AI text extraction solutions.

BlueVine Inc., a heavily funded California startup with an online banking platform for businesses, used Textract to process Paycheck Protection Program loan applications. According to AWS, the startup managed to build an automated loan processing workflow in a few days with Textract at its foundation. BlueVine processed more than $4.5 billion worth of loans for about 155,000 small businesses in four months and estimates that its efforts contributed to the saving of 470,000 jobs.

Electronic signature provider HelloSign is another AWS customer that has adopted Textract to help it meet increased demand. HelloSign, a Dropbox Inc. subsidiary as of last year, used Textract’s document parsing features to let users of its HelloWorks service process forms and other paperwork with less manual work. HelloWorks helped an unnamed “popular on-demand grocery delivery service” onboard millions of shoppers in a few weeks, while a New York-based urgent care adopted the service to register new patients.

AWS is also supporting efforts to develop better COVID-19 forecasting models. The cloud giant detailed today how a research group comprising the University of California at Irvine, Rackspace Technology and others has built a model on its platform that can generate disease forecasts for countries, states, counties and specific ZIP codes. The group used Amazon Forecast and AWS’ DeepAR+ algorithm in the project. 

Amazon Forecast is a service for running AI models on time-series information, or collections of data points organized in the order they were created. The DeepAR+ algorithm, in turn, is a pre-packaged supervised machine learning algorithm for generating forecasts from time-series information.

The AI feature set that AWS has built out in its cloud is finding use across a broad range of areas. Carrier Global Corp. is using AWS’ AI tools to power Lynx, a recently announced analytics platform it’s developing to assist companies such as retailers with optimizing their supply chains. The NFL, meanwhile, last year partnered with the cloud giant to improve player health and safety using machine learning. 

Image: AWS