Full Stack Developer - Manager

Role purpose You'll build cloud-native services and user interfaces for a multi-year programme modernizing a large portfolio of middle-office applications and moving them to AWS. This is a hands-on development role with a strong back-end focus. You'll design and build Java microservices and APIs on AWS, and also contribute to ReactJS/TypeScript front ends. You'll use AI coding tools such as GitHub Copilot, Claude Code and Devin every day to deliver faster, while staying fully accountable for the quality and security of the code. Key responsibilities Back-end development (primary) Design and build Java/Spring Boot microservices and REST APIs for trade processing, accounting, reconciliation and settlement functions. Build event-driven integrations with Kafka and IBM MQ, including message design, error handling, retries and idempotency. Write efficient data access against Oracle (JPA/Hibernate and SQL), working with database engineers on schema design and performance. Rebuild legacy business logic as clean, testable services, working with business analysts and subject-matter experts to confirm behaviour. Write unit, integration and contract tests, and help with performance and resilience testing. AWS Build and deploy services on AWS: ECS/EKS, Lambda, API Gateway, S3, SQS/SNS, RDS, CloudWatch and IAM. Design services for scalability, high availability and security, following AWS Well-Architected principles. Work with DevOps on Terraform, Docker and CI/CD pipelines (Harness/GitHub Actions) so your services deploy automatically. Front-end development (secondary) Build and maintain ReactJS/TypeScript screens and components for data-heavy enterprise applications that replace legacy desktop UIs. Connect front ends to back-end APIs, handling authentication (SSO/MFA), state management and error handling. Write maintainable JavaScript/TypeScript and component tests (Jest/React Testing Library). AI-assisted development Use GitHub Copilot for inline code completion, chat-based refactoring, test generation and pull-request summaries. Use Claude Code for agentic, multi-file work in the terminal: understanding unfamiliar codebases, refactoring across modules, generating tests and fixing failing builds. Work with Devin (or similar autonomous agents) on well-defined tasks: write clear task specs, review its pull requests and correct its output. Use AI to analyse legacy code (Java EE, Sybase SQL, PowerBuilder, COBOL) and draft specs for subject-matter experts to confirm. Write good prompts and context files (e.g. README files, coding standards, agent instruction files) so AI tools produce code that follows team standards. Review all AI-generated code for correctness, security, performance and licence compliance before merging. You stay accountable for every line. Share effective prompts, patterns and lessons learned with the team. Ways of working Take part in code reviews, design discussions and agile ceremonies.