Home / Projects / Fixed Income Simulator Fixed Income Simulator Emirates NBD · Sep 2025 - Present · web
Internal wealth management platform enabling Relationship Managers and Wealth Advisors to simulate and generate fixed-income portfolios across a defined securities universe. Users select securities, enter investment amounts subject to configurable validation rules, and receive multi-scenario analytics with graphical insights across interest, notional distribution, return on investment, and cashflow projections.
Product Development & Simulation Engine Delivered end-to-end product implementation from business requirement through production release Built securities selection workflow with configurable investment validation rules across selected holdings Implemented portfolio simulation engine generating multiple financial scenarios with interactive graphical outputs:Interest calculation and accrual projections Notional distribution breakdowns Return on investment (ROI) analysis Cashflow forecasting and timeline views Developed REST APIs and MongoDB data models to persist securities, simulation inputs, and computed portfolio results AI Agent & RAG Architecture Designed and built a conversational AI agent using RAG to assist advisors within the simulator workflow Implemented multi-node agent orchestration with intent classification routing requests to specialized handlers:Search handler for contextual data retrieval Calculation handlers for financial computation and scenario analysis Report handler for structured result generation and response formatting Integrated LLM-based reasoning with dynamic chatbot rendering based on response format (text, tables, charts, and structured reports) Applied token optimization, context engineering, and reusable agent skills to improve accuracy, latency, and cost efficiency Frontend Engineering Built advisor-facing UI using React.js and TypeScript with modular flows for selection, investment input, and scenario comparison Developed reusable chart and visualization components for simulation outputs and agent-driven report display Implemented chatbot interface with format-aware rendering for agent responses across multiple result types Testing & Code Quality Authored end-to-end, unit, and integration test suites covering both backend services and frontend workflows Implemented automated test coverage with Jest and enforced code quality gates through SonarQube analysis Validated business rules, simulation logic, and agent behaviour through targeted integration scenarios AI-Assisted Development & DevOps Leveraged AI coding assistants for day-to-day feature development, debugging, and iterative delivery across the full stack Applied context engineering practices to structure prompts, skills, and agent workflows for maintainable AI features Containerized application services with Docker and deployed via OpenShift with Jenkins CI/CD pipelines Technologies JavaScript React.js TypeScript Node.js Rest APIs MongoDB GPT-4 RAG Agent Orchestration Claude Docker OpenShift Jenkins Jest SonarQube Git Github LangChain / LangGraph