Accelerating Healthcare Technology Innovation Using Amazon Q and Amazon Bedrock with Availity

Availity operates an encrypted solution that connects medical providers with health plans across the United States, serving as a critical intermediary in healthcare. The healthcare industry is traditionally slow to adopt new technology, but Availity has taken a progressive approach.

“As Availity has grown over the past 20 years, it has become increasingly important for us to select where we spend our engineering time,” says Privat. “We need to spend our time designing solutions for our customers rapidly. Essentially, we need to supercharge our development staff.”

Availity’s software development teams were getting bogged down by time-consuming but necessary tasks. Engineers spent hours sitting in coordination meetings, filing documentation, and conducting routine assessments instead of solving complex issues. The standard scrum model required time for security reviews, cost evaluations, and performance testing—critical requirements in healthcare that nonetheless created significant overhead. For example, a typical release review would require a 3-hour meeting where senior engineering leaders manually assessed hundreds of release packages, examining each for potential risks.

Availity saw an opportunity to transform its development process with AI. In January 2024, the company began exploring the capabilities of Amazon Q. Availity first adopted Amazon Q Developer, a generative AI–powered assistant for software development, as a code completion tool.

After deploying Amazon Q Developer to a few engineers, Availity saw that there were also use cases beyond code completion. Availity’s developers use the embedded chat to pair programs through complex solutions; typically, a developer suggests a few options, and then Amazon Q Developer provides the implementation details. In one example, an engineer resolved an issue in one 30-minute interaction with Amazon Q Developer rather than the week it would have taken previously.

Because all Availity’s data was stored on AWS, the company quickly realized that it could use Amazon Q Developer to streamline many other development tasks, such as writing documentation and conducting security assessments. Availity developed specialized bots using Amazon Bedrock and Amazon Q Business, a highly capable generative AI–powered assistant for finding information, gaining insight, and taking action at work. These bots are directly incorporated into Availity’s development tools and chat systems, helping engineers brainstorm solutions and generate code.

For example, the Production Readiness Bot monitors development progress and guides teams through the process of taking code to production, automating documentation requirements and highlighting potential issues early in development. The Risk Assessment Bot analyzes release packages for security considerations, and the Security Bot conducts automated scans and suggests remediation steps. These bots proactively flag requirements early in the development cycle, helping teams complete critical tasks, such as creating runbooks, weeks in advance rather than days before release.

“I’ve witnessed engineers having successful brainstorming discussions using Amazon Q Developer,” says Privat. “Having that conversation probably saved us 3–4 days’ worth of work. Amazon Q is completely different from just having an AI chatbot or an AI assistant in your integrated development environment.”

Since implementation, Amazon Q Developer has generated 33 percent of all code written by the engineers who use it. These development teams have also accepted 31 percent of the tool’s suggestions—resulting in 19,500 lines of AI-generated code. The system has automated traditionally time-consuming tasks such as code testing, security assessment, and performance testing, which engineers often cite as the most tedious parts of their work. Using the Risk Assessment Bot, Availity has autonomously conducted 12,600 security scans, analyzed code, and summarized findings. Engineers can directly ask the bot about specific security risks in their code changes and receive immediate insights about potential concerns. Thus, teams can review releases in minutes instead of hours, focusing more on improving security posture and performance.

Availity’s approach to AI incorporation has changed how its engineers work. Rather than simply using AI for code completion, the company has created an environment where AI acts as an intelligent programming partner, handling routine knowledge work so that engineers can focus on innovation.

Having established this AI framework for development teams, Availity is now exploring similar applications for other internal teams. The company is developing specialized bots to assist data engineers and data scientists with their workflows. Plans are also underway to create AI-powered solutions for sales and customer service teams, including a specialized bot that helps map healthcare networks and analyze customer relationships.

“We established two distinct tracks for this project: (1) improving the developer environment and (2) creating specialized process automation bots,” says Privat. “The AWS team has delivered on track (1), and my team has delivered on track (2). This makes for great results.”