Jerin Joseph

A multi-tenant membership platform for AICPA and CIMA, serving 600K+ members. The architecture is a hybrid — microservices, event-driven messaging, and deep integrations with third-party SaaS services covering CRM, CMS, billing, product management, and authentication. The kind of system where understanding how everything connects matters as much as writing the code.

• ▸Designed and delivered full-stack features for a multi-tenant B2B and B2C membership platform (AICPA/CIMA) serving 600K+ members, built on a serverless, event-driven microservices architecture spanning 9+ integrated systems, including CRM, billing, e-commerce, CMS, and authentication.
• ▸Protected £157,620+ in revenue and resolved data integrity issues across 385,000+ records spanning member organisations, employee accounts, and exam synchronisation workflows through cross-system debugging and root-cause analysis. Collaborated with product, infrastructure, and third-party integration teams to deliver resilient, long-term solutions.
• ▸Recovered a lost Zuora–CommerceTools rate plan integration by redesigning it as an automated event-driven pipeline using Amazon S3, AWS Lambda, and DynamoDB. Eliminated 2,400+ hardcoded records, unblocked new product launches (including the organisation's AI offering, JOSI), and resolved a compliance risk that exposed the business to potential regulatory fines.
• ▸Improved Assessor Dashboard load times by 93% by diagnosing and resolving N+1 data fetching, redundant frontend API calls, and asynchronous processing inefficiencies through request batching, parallelisation, and a 50% reduction in API payload size.
• ▸Led the proof of concept for AI-assisted development using Claude Code, defining shared context architecture, configuration standards, and engineering workflows to improve developer productivity and consistency.
• ▸Modernised the testing strategy by introducing Vitest, reducing unit test execution times by up to 80% and improving developer feedback loops across local development and CI/CD pipelines.
• ▸Led weekly sprint release cycles as the lead developer within an Agile/Scrum environment, while mentoring engineers through pair programming, onboarding documentation, and knowledge-sharing workshops.

Two years in the transportation industry, spanning both product engineering and embedded systems — full-stack work on React frontends and Java APIs, alongside real-time software for onboard computers and event recorders operating under strict reliability constraints.

• ▸Rebuilt a high-traffic application dashboard by consolidating data from a legacy system API and a Java service into a unified backend API. Migrated calculation logic from the frontend to the backend and implemented client-side optimisations, including memoisation and lazy loading, reducing load times by 45%.
• ▸Built an end-to-end commenting feature, developing the Spring Boot REST API and implementing an optimistic UI update pattern in React that updated local state without reloading the dashboard, with robust error handling and rollback for failed saves.
• ▸Automated the manual build and deployment process for onboard software by developing a Bash-based tool that handled compilation, packaging, upload, and on-device loading with comprehensive logging, reducing build cycle time by 15 minutes.
• ▸Developed real-time software modules for locomotive event data recorder systems, extending event-capture capabilities to record additional operational data while meeting stringent safety, reliability, and regulatory requirements.

A large-scale warehouse management system built on Java, PL/SQL, and Oracle DB, supporting 8 geographically distributed U.S. locations. High-volume and mission-critical — the kind of environment where careful root cause analysis matters more than moving fast.

• ▸Developed and maintained a large-scale warehouse management system built with Java, PL/SQL, and Oracle Database, supporting mission-critical logistics operations across eight geographically distributed U.S. locations.
• ▸Led root-cause analysis and resolution of production incidents, implementing permanent fixes that reduced recurring issues and support tickets by 25%. Recognised with the Infosys Insta Award for demonstrating strong ownership and high-impact delivery.
• ▸Redesigned an internal IBM MQ message reprocessing tool to support multiple queue formats and automated csv data ingestion, eliminating manual intervention for failed message reprocessing and improving the resilience of distributed messaging workflows.