Staff Software Engineer, AWS Platform

*Please note this is a hybrid position located in Buffalo, NY. Candidates should live in/around towns/cities that are within a reasonable commutable distance.*

Tech stack: Go (Golang), AWS Lambda, DynamoDB, PostgreSQL/Aurora, EventBridge, SQS, Step Functions, AWS SAM, serverless microservices, CI/CD, React and React Native.

Valmar Merchant Services is building and operating the infrastructure behind a high-velocity payments platform. Our environment is not theoretical: approximately 14 Go (Golang) microservices, 129 AWS Lambdas, DynamoDB as the system of record, Aurora for reporting, EventBridge, SQS and Step Functions across asynchronous workflows, and six-plus environments promoted through SAM.

We are looking for a Staff Software Engineer who has operated this kind of AWS serverless estate in production - through incidents, deployments, failures and multi-environment promotion - and knows how to keep a complex platform coherent while a small team continues to ship. This is a hands-on player-coach role. You will set technical standards and make architecture decisions, but you will also write the hard path yourself.

Payments or ACH experience is a plus, not a prerequisite. We can teach the payments domain. We are not looking to teach the fundamentals of Lambda, DynamoDB, IAM or event-driven architecture on AWS at this level. The right person brings deep platform judgment, ownership under pressure, and the ability to make the engineers around them faster without adding unnecessary process or bureaucracy.

What You Will Own

• Own the AWS platform seams: Lambda concurrency and timeouts, DynamoDB streams, EventBridge versus synchronous cross-service calls, IAM and environment-variable coupling, and patterns for work that exceeds the Lambda execution limit using Step Functions, Fargate or the appropriate alternative.
• Keep multi-environment delivery reliable across nested SAM stacks, personal development stacks, LocalStack, and shared QA, UAT and production environments, including versioned schema migrations after code is live. Go 1.24 services deploy through SAM using a fully scripted build-and-deploy loop.
• Set and protect the dual-store architecture: DynamoDB as the source of truth and PostgreSQL/Aurora as the reporting read model, including replay strategy, stream-consumer failure handling and lag SLOs.
• Own the authorization seam across API Gateway authorizers, Cognito and our custom RBAC model. Every new API endpoint must ship with audit logging and authorization wiring from day one.
• Raise the operating floor with runbooks, actionable alarms, written IAM/EventBridge standards, compatibility policies for the public API and generated SDKs, and a paging path the team can trust.
• Write production code and carry difficult features across the stack when needed. Our frontends are React and React Native, and platform changes may include UI work.
• Make pragmatic modernization decisions. New backend work lands in a standardized service layout while much of the production code remains in legacy trees. You will help determine the right conversion order rather than defaulting to a greenfield rewrite.
• Share ownership of CI/CD, deployment safety and the paging path for critical batch, settlement and synchronization workloads. When something breaks, you are comfortable diagnosing, stabilizing and improving it.

What Success Looks Like

• The platform becomes easier to operate as it grows: fewer fragile seams, clearer standards, better observability and less deployment risk.
• Engineers can move quickly without sacrificing authorization, auditability, data integrity or operational discipline.
• Incidents produce better systems. Runbooks improve, alarms become more actionable, failure modes are understood and repeat problems are engineered out.
• You make the team faster by pairing strong technical judgment with hands-on execution - not by creating layers of review.
• Architecture decisions are pragmatic and grounded in the system we actually operate, not the platform you would prefer to rebuild from scratch.

What You Bring

Expect to be interviewed on the core requirements below.

• Production AWS serverless at meaningful scale - beyond a handful of Lambdas behind API Gateway. You have owned IAM, aliases, concurrency, throttles, cold starts versus reserved capacity, and deployment blast radius across multiple environments.
• DynamoDB as a system of record, including keys, GSIs, streams, …