BackEnd Software Engineer

Reports To: CTO

Department: Engineering

Position Type: Independent Contractor (Fractional)

Travel Expectations: 1-2 days/week onsite in Saint Paul, MN

Location: Saint Paul, MN (Hybrid)

Date: December 2025

We're building something special for the TTRPG community—a platform that understands the magic of bringing people together around the table. This fractional contractor position is responsible for implementing the backend services, APIs, and infrastructure that power The Keep's event platform. You'll build the resilient, scalable systems that handle everything from user authentication to real‑time event processing, ensuring game masters and organizers can focus on what matters: creating unforgettable experiences for their communities.

We have an aggressive roadmap ahead of us. You'll work directly with the CTO and collaborate with other engineers, Product Management, and UX Designer(s) to ship features weekly—if not daily. This is a high‑autonomy, fast‑paced environment where your work powers real events for real people. If you love building elegant, event‑driven systems and thrive on seeing your infrastructure enable (tens of) thousands of gaming sessions, this is your table.

You'll be working with modern, production‑grade technologies:

• Language: Go
• AWS Services: Lambda, API Gateway, Dynamo
  
  DB, EC2, ECS, S3, Cognito
• Event Streaming: Apache Kafka (self‑managed on EC2)
• Infrastructure: Terraform (100% IaC. Everything is code and automated)
• Observability: LGTM stack (Loki, Grafana, Tempo, Mimir)
• Deployment: Automated CI/CD pipeline
• Collaboration: Git workflows with automated testing gates before deployment

01 | API Development and Implementation: Design and implement RESTful APIs using AWS Lambda and API Gateway, following established architectural patterns for authentication, validation, error handling, and response formatting. Success is frontend developers shipping features confidently, users experiencing instant responsiveness, and zero API‑related production incidents.

02 | Microservices Development: Build and maintain serverless microservices that handle discrete business logic domains within our event‑driven architecture. Success is services running reliably at scale, clear domain boundaries that accelerate development, and minimal cross‑service coupling.

03 | Event Consumer Implementation: Develop Kafka event consumers that process asynchronous events, update data stores, trigger workflows, and maintain system consistency across services. Success is zero data loss, sub‑second event processing, proper error recovery, and the ability to replay events to reconstruct system state.

04 | Data Modeling and Database Management: Design and implement Dynamo

DB data models optimized for access patterns, ensuring efficient queries and appropriate use of indexes. Success is consistently fast queries regardless of data volume, efficient cost scaling, and data models that evolve cleanly with new features.

05 | Infrastructure as Code: Maintain and extend infrastructure definitions using Terraform in our fully automated, 100% IaC environment where all changes are made through code—humans do not have direct access to control planes. Success is deployment reliability, infrastructure that documents itself, seamless environment provisioning, and zero manual infrastructure changes.

01 | Dev Ops and CI/CD: Contribute to continuous integration and deployment pipelines, ensuring automated testing, builds, and deployments across multiple repositories. Success is pipelines that just work, deploying multiple times daily with confidence, and reduced time from commit to production.

02 | Monitoring and Observability: Implement logging, metrics, and tracing using the LGTM stack to enable proactive issue detection and debugging. Success is catching issues before users notice them, diagnosing problems in minutes instead of hours, and having the data to optimize performance continuously.

03 | Container Management: Maintain and deploy containerized services using ECS where appropriate for workloads that don't fit the serverless model. Success is containers…