James Nixon

Led an org-wide AI-assisted development practice, defining how engineers ground AI tooling in real codebase context to produce reviewable, auditable output across multiple teams. Delivered early end-to-end mortgage decisioning capabilities in a polyglot TypeScript/Go environment, including data ingestion, rules evaluation, structured output, and auditable logging.

Owned backend delivery end-to-end (design → build → release → run) across TypeScript/Node, NestJS, and microservices. Led a 2M+ record Strangler-pattern migration and built AWS Neptune graph-backed recommendations over 9M+ data points, improving relevance from production signals.

Re-architected key APIs from synchronous to event-driven using EventBridge and SQS, cutting response times from ~3–4s to <100ms. Built an event-driven email notifications pipeline and hardened async processing with idempotency, retries, DLQs, and monitoring while managing Terraform and Kubernetes infrastructure.

Built and maintained authentication APIs in TypeScript/Node to support login/account flows for millions of users. Delivered Jagex Launcher UI features in React and Next.js and set up CI/CD for safe, repeatable launcher deployments.

Led backend development from zero to launch, designing and building core services and APIs for creator discovery, campaigns, and data ingestion. Integrated TikTok APIs to ingest and normalize creator data with robust rate-limit/failure handling and scaled the platform to 100,000+ users while running backend hiring and setting engineering standards.

Built customer-facing features with React and backend services using TypeScript/Node and NestJS, introducing event-driven architecture with RabbitMQ to offload heavy workflows. Led a GitLab → GitHub migration and owned Docker + Kubernetes deployments and CI/CD pipelines, contributing to technical hiring.

Maintained a production Perl monolith while building new TypeScript/Node microservices to modernize and decouple the platform. Owned CI/CD (GoCD, Jenkins, GitHub Actions) supporting 2–3 production deployments per day and built GCP + Terraform infrastructure while participating in on-call and incident response.