Kalim Ullah Khan

Led architecture rewrite of the AI-powered education widget app, defined scalability strategy, and enabled curriculum-aligned AI feature expansion. Established modern frontend foundation (Vite, React, TypeScript), defined project standards and governance, and built AI-assisted teacher companion workflows (lesson plans, knowledge checks, assignments, math practice) to boost teacher productivity. Reduced bundle size by ~30% and improved load performance by ~25% through dependency optimization and architectural refactoring (validated using New Relic metrics). Leveraged analytics to track feature adoption and optimize AI tutoring workflows.

Led frontend foundation for an AI-powered education wallet app and built AI-assisted teacher workflows (lesson plans, knowledge checks, assignments, math practice). Optimized performance by reducing bundle size ~30% and improving load times ~25% via dependency optimization and architectural refactoring. Established analytics to track feature adoption and inform AI tutoring workflows.

Board Observer – an AI-powered meetings app. Implemented features with Next.js, Vercel AI SDK, Python FastAPI, and ChatGPT-4. Built auto meeting summarizer generating concise AI summaries from transcripts, enabled live document sharing with per-page AI insights, and integrated a RAG-powered contextual Q&A assistant. Optimized chat performance using React Window and collaborated with backend to surface real-time data.