AWS Public Sector Blog

Precise Software Solutions implements ML as a service on AWS to save time and money for federal agency

AWS branded background design with text overlay that says "Precise Software Solutions implements ML as a service on AWS to save time and money for federal agency"

Precise Software Solutions, Inc. (Precise), an Amazon Web Services (AWS) Partner, participated in the AWS Think Big for Small Business Program (TBSB) to expand their AWS capabilities and to grow their business in the public sector. After completion of the program, Precise achieved Advanced tier partner status and was selected by a federal government agency to create a machine learning as a service (MLaaS) platform on AWS. The platform helped the agency digitize and process forms, pictures, and other documents. Using the platform, which uses Amazon Textract, AWS Fargate, and other services, the agency gained a four-fold productivity improvement by streamlining and automating labor-intensive manual processes.

Expanding a solid partnership

Precise is a certified small business and SBA 8(a) program participant that provides services and solutions that help federal, state, and local government organizations enhance and expand their IT capabilities. The TBSB program offers small and minority-owned public sector organizations access to business, technical, and marketing support from AWS. By participating in the program and achieving Advanced tier status, Precise was able to expand their business.

“As an AWS Partner and a participant in the Think Big for Small Business, we’ve received solid support from AWS, which has enabled us to build innovative solutions for our customers,” says Ben Duan, Chief Technology Officer at Precise.

The federal government agency Precise worked with needed to automate manual processes for document intake and image processing. “The agency wanted to use AI [artificial intelligence] and ML to automate document digitization, and it also needed help understanding each document it digitizes,” says Duan. “For image processing, the agency does a lot of inspections and takes a lot of pictures. This customer wanted to use machine learning as a tool to digitize images and recognize handwriting.”

Helping government agencies adopt AI and ML technologies

Precise works closely with AWS to offer end-to-end cloud services such as enterprise cloud strategy, infrastructure design, cloud-native application development, modern data warehouses and data lakes, AI and ML, cloud migration, and operational support.

The demand for modernization is growing, and Precise can help government agencies adopt AI/ML technologies. “A lot of federal agencies know machine learning will be beneficial, and need partners to augment the expertise, help consolidate siloed data, and enhance the capabilities in their existing environment,” says Duan.

Creating an MLaaS solution on AWS

Precise worked with the federal government agency to evaluate their needs and established a team of solution architects, software developers, data scientists, and cloud engineers to create an agency-wide MLaaS platform. The MLaaS platform contains pre-built models and algorithms that address the government agency’s most critical use cases by using computer vision, image classification, natural language processing (NLP), and other ML technologies.

Built on AWS, the platform offers common AI/ML services such as handwriting recognition, foreign language labels translation, form recognition, image recognition, content extraction, document understanding, and workflow automation. The solution uses a microservices architecture based on Amazon Elastic Container Registry (Amazon ECR) and uses AWS serverless technologies, including AWS Fargate, for flexible and scalable compute capabilities. The solution also relies on the Amazon Textract ML service to automatically extract, translate, and understand text, handwriting, and data from scanned documents.

Precise also worked with the agency to obtain an Authorization to Operate (ATO) for security clearance to use the MLaaS platform.

“It was critical for the government agency to obtain an ATO for security requirements, and we helped them achieve this two months faster than normal because the AWS based container environment enabled us to speed up the security control implementation,” says Wilbur Chen, the lead architect for the MLaaS platform.

Saving time and money by automating manual processes

Using the MLaaS platform, the federal government agency automated key business processes. “With our MLaaS solution on AWS, we helped the agency digitize and automate its previously manual tasks for processing large volumes of forms and data,” Chen says.

As an example, the agency is using the solution to automatically correct documents to comply with Section 508 of the Rehabilitation Act of 1973, which requires agencies to make electronic information accessible to people with disabilities. For compliance, the agency must review and, in some cases, update each page of a document to make sure all text aligns with accessibility requirements.

“The MLaaS platform automates most corrections, reducing the average time it takes to correct a page from 20 minutes to less than five minutes—less than one minute for ML autocorrection and four minutes for human review,” Duan says. “That also translates to a labor cost savings, so the agency reduces the cost to do a full document review by 80 percent.”

Sharing data, sharing costs

Different business units within the government agency are taking advantage of the fact that MLaaS provides a common, shared platform. As a result, data and solution silos are eliminated, and infrastructure and licensing costs are shared by all departments that consume the platform’s services.

“With a lot of machine learning solutions, each department has to pay to stand up and use each service on its own,” Duan says. “This solution gives the entire agency an enterprise-ready shared platform for all departments to consume from an API service or UI.”

The MLaaS platform is designed to be easy to adopt, and Precise plans to offer it to additional government agencies in the future. “There are a lot of other use cases that can use this same model for consuming enterprise services,” says Duan.

Small business, big impact

Participating in the Think Big for Small Business program, Precise strengthened its partnership with AWS and achieved Advanced tier status. This collaboration empowered Precise to develop a transformative MLaaS platform for the federal agency.

With AWS support and Think Big for Small Business program benefits, Precise remains at the forefront of AI and ML innovation in the government sector

Interested in transforming your business? Discover how you can join the Think Big for Small Business program and drive your company’s success.

Jonathan Cooper

Jonathan Cooper

Jonathan is the global lead for the Think Big for Small Business Partner Program at Amazon Web Services (AWS). He represents AWS to the public sector, serving small and diverse-owned businesses around the world that are seeking to grow their partnership with AWS. Jonathan works to develop the AWS capabilities of these businesses, enabling them to better meet the evolving needs of public sector customers.

Allison Guse

Allison Guse

Allison Guse is a partner marketing program manager at Amazon Web Services (AWS) for the Worldwide Public Sector Partner Marketing team. She manages the AWS Partner reference program to promote partners, their services, and solutions globally, targeting government, education, healthcare, and nonprofit audiences.