Case study: Modernizing Microsoft-Stack to Cloud Native & License Free on AWS
Overview
Hirevue is a U.S.-based HR technology company rooted in the science of skill validation and data-driven hiring. Their AI-powered solutions drive smarter hiring, predict performance, and boost ROI, empowering employers to find quality talent at speed and scale. Their platform, Talent 2 Opportunity (T2O), is Hirevue’s vision for the future of hiring, and is the most comprehensive talent acquisition platform on the market.
T2O is built largely on .NET with C# and runs on fleets of EC2 Windows servers backed by AWS-managed RDS SQL Server in multiple AWS regions serving customers globally. Hirevue partnered with Evolve Cloud Services to modernize T2O on AWS to achieve FedRAMP compliance while dramatically reducing costs, and improving scalability, deployment times and performance.
Challenge
Hirevue needed to modernize their Talent 2 Opportunity (T2O) platform running on AWS. Core drivers for modernization included achieving FedRAMP compliance, lowering AWS costs, improving performance, and enabling teams to independently deploy and scale individual T2O services.
Meeting these goals wasn’t feasible with the existing Windows and SQL Server tech stack and monolithic application deployment pattern. As the only FedRAMP-authorized hiring solution for the public sector, the certification for T2O was a key objective, and hardened Linux containers operating in EKS offered a more streamlined path for FedRAMP certification than legacy Windows infrastructure. Hirevue also needed to reduce costs in order to continue scaling out their cloud infrastructure to meet customer growth needs. Hirevue chose to partner with Evolve Cloud Services to support modernizing T2O.
Evolve delivered AWS-funded Modernization Viability Assessments (MVAs) to dive deep into the infrastructure, codebase, and database schema of T2O’s 60+ services and 20+ databases. Working with HireVue’s tech team and AWS solutions architects, Evolve defined a target architecture, database platform and tech stack based on Linux containers in EKS and Amazon Aurora PostgreSQL with Babelfish.
The MVA:
- Identified Hirevue’s modernization goals
- Analyzed source code and dependencies to identify Windows-specific code and packages
- Assessed 20+ database schemas and documented specific schema and code changes required for migration to Amazon Aurora
- Provided a details cost-savings analysis for the target platform
- Delivered a proof-of-concept with one T2O service running on Linux against PostgreSQL both with and without Babelfish
- Proposed a modernization project with detailed plan and level of effort
Solution
AWS and Evolve worked with Hirevue to kick off the modernization effort with an on-site build-a-thon at Hirevue’s headquarters, which allowed the technical teams to jointly solve technical challenges to reduce risk and validate architectural designs. The overall effort was broken into two projects: one to move the T2O .NET stack to EKS Linux containers, and the second to migrate the databases to Amazon Aurora. This allowed Hirevue to begin their FedRAMP certification only six months after the build-a-thon, with the move to Aurora PostgreSQL with Babelfish as a “fast follow”, and initial databases moving to production in early 2026.
Modernizing a platform as vast and complex as T2O—transitioning over 60 services and 20 databases from a legacy Windows stack to Linux containers and Amazon Aurora—is a feat few would even attempt. It required a complete overhaul of our architecture, but our collaboration with Evolve Cloud Services and AWS allowed us to achieve what many thought was unreachable. We truly made the impossible, possible, securing our FedRAMP certification while building a foundation for unprecedented scale.
— Alberto Silveira, CTO, Hirevue
.NET to EKS Linux Containers
Evolve iteratively reviewed, refactored, and tested more than 60 T2O services to remove dependencies on Windows Server in the codebase, configuration and binaries, and to replace synchronous calls with a more modern and scalable asynchronous pattern. New Docker and Helm configurations were implemented for deployment across environments including AWS GovCloud and networking was re-implemented using Kubernetes networking, Istio, and AWS ALBs. Other substantial refactoring included rewriting a scheduled-task processor to eliminate 80% of the code, and converting .NET Framework Windows Services to containerized .NET 8 services. Evolve also supported the HireVue’s DevEx team in the creation of new CI/CD pipelines based on GitHub. Additional architecture improvements followed, including the optimization of the container build pipeline and the Docker layer optimization to maximize image caching. Evolve also prototyped a local and cross-platform container-based dev stack for Hirevue developers on Windows or macOS.
RDS SQL Server to Amazon Aurora PostgreSQL
One of the most expensive components of any large workload in the cloud is licensed databases, which are often up to two-thirds of customers’ total cloud bill for a data-heavy application. Evolve analyzed the T2O codebase to determine the service-database usage, and worked with Hirevue to determine the order for migration to Aurora, starting with usage patterns and progressing to more complex cases.
For each database, Evolve worked closely with the Hirevue cloud engineering team to provision databases using Hirevue’s IaC platform and then tested each T2O service against Amazon Aurora, making code changes where required. Hirevue’s own QA team then regression-tested services to validate success. This iterative and collaborative approach allowed us to complete the migration of all the databases to Amazon Aurora in the Dev environment in under 90 days, with the first databases planned for production less than one month later.
Early Results & Next Steps
Early tests with the modernized T2O services running in EKS showed promising performance gains, and helped Hirevue complete their FedRAMP certification for AWS GovCloud. Additional benefits are that individual services can now be deployed and scaled independently allowing Hirevue to make more efficient use of their cloud resources, in addition to the roughly 30% – 40% cost savings from moving off of EC2 Windows. The move to Amazon Aurora PostgreSQL is anticipated to save more than 50% of T2O’s annual RDS database costs. Amazon Aurora PostgreSQL can also scale to three times more read replicas than RDS SQL Server and supports advanced features like Global Database and serverless configuration. Both the T2O service modernization to EKS and containers and the database migration to Amazon Aurora were completed on time and within budget.
Hirevue plans to achieve even more price-performance gains for T2O by testing and ultimately deploying services to AWS Graviton arm64 compute, on which the Amazon Aurora database clusters already run. This allows Hirevue to handle more customers and planned growth without spending more on infrastructure.