At LexiAssist, I built and scaled an educational platform across Next.js web and React Native mobile clients. I implemented WebSockets and Server-Sent Events for real-time sync and live AI audio transcription.
On AI Platform Optimization, I audited a FastAPI/LangGraph integration, cut API request volume 50–100x by batching embedding calls, and integrated a Weaviate RAG pipeline that reduced token consumption up to 30x. I also hardened a Go gateway with validation and rate limiting, and added a Redis cache to prevent duplicate document processing.
