Arjun Ambavane

I build scalable, production-grade full-stack systems with a focus on real-time performance and AI-enabled capabilities. My work centers on turning complex requirements into reliable features—fast, testable, and deployable.

In a recent Full Stack SDE Intern role, I built an AI-driven cost-estimation system for 3D models, supporting 500+ MB files with a scalable tRPC backend. I also implemented semantic material-matching using VectorDB + LLM retrieval across 5,000+ cost-sheet entries, and deployed a fully containerized production platform using MinIO for large-file storage and retrieval.

Earlier, I engineered a research showcase portal with Next.js and strengthened CI reliability. I replaced manual QA with a Playwright-based testing pipeline on AWS, auto-triggered on PR merges, reducing testing time by 80%.

I also developed a real-time GUI for a safety-critical Hyperloop prototype, processing 50+ telemetry events/sec with <40ms latency using React.js, Node.js, WebSocket, and MongoDB. Alongside core engineering work, my team represented MIT WPU at European Hyperloop Week 2024 in Zurich and secured 2nd place in the Mechanical & Electrical Subsystems category.