Kelvin Smith

Developed a dynamic and responsive UI for the AI chatbot using React.js, integrated Redux for global state management, built reusable components for chat windows, input fields, and message displays, leveraged React Hooks for state and side effects, and implemented lazy loading/code-splitting to improve load times during peak traffic.

Led the design and architecture of a production-ready serverless, event-driven platform that processes real-time data, ensuring system reliability and high scalability. Built AI systems leveraging orchestration frameworks like Mastra to integrate multiple AI models, enabling seamless task routing between generation, classification, and structured reasoning models. Engineered data pipelines to process webhooks, CRM data, and other event-driven signals, transforming raw input into structured insights for actionable decision-making. Developed and scaled retrieval systems that combined proprietary knowledge bases with context-aware AI generation, optimizing the recommendation engine’s accuracy. Applied a reliability-first mindset, ensuring high uptime and low latency through robust failure isolation, durable queues, and retry mechanisms.

Led the development of a Next.js-based platform for real-time cardiac signal analysis, optimized page load times by reducing JavaScript payloads and unnecessary re-renders, implemented dynamic routing and API routes with REST APIs, integrated MySQL with efficient queries and indexing, and enhanced UI with Tailwind CSS and Framer Motion; deployed on Vercel.