Rajkumar Murugesan

Leveraged LangChain and LangGraph to build stateful AI agents that power adaptive mock assessments, dynamically adjusting question difficulty based on real-time candidate performance. Developed a B2B scoring algorithm providing detailed analytics on candidate strengths and weaknesses, backed by a PostgreSQL Aurora data layer for high availability. Secured the platform with RBAC and API rate limiting, deployed on AWS ECS/ECR behind an Application Load Balancer and WAF for DDoS protection.

Engineered a system of 4 decoupled microservices and developed AI agents using OpenAI's function calling to automate complex marketing workflows, including a customer segmentation engine that analyzes purchase history to generate high-conversion personalized email campaigns. Architected a plug-and-play Shopify integration for real-time store data sync and leveraged Redis queuing to handle burst traffic during campaign dispatch. Managed the full AWS deployment lifecycle using EC2, SES, and S3, with AWS Cognito for enterprise-grade OAuth and JWT-based session handling.

Developed robust backend logic to manage meeting room bookings, resolving overlapping schedules and resource conflicts through atomic database transactions. Built a CRUD system for workspace managers to track and assign amenities to specific bookings, and designed an RBAC system for corporate admins to control permissions across employees, visitors, and facility managers. Delivered features within a high-velocity two-person Scrum team on bi-weekly sprint cycles.

Led the complete architectural revamp from a legacy framework to NestJS, restructured the PostgreSQL schema to reduce query latency by over 40%. Integrated Razorpay with secure webhook handling for order confirmations and refunds, alongside automated SMS and email notifications for order updates and rental expiration alerts. Applied OOP principles throughout the backend to deliver a modular, testable, and maintainable codebase.