Neeraj Rajeev Shetkar
Building Intelligent Interfaces
Bridging Large Language Models with On-Device Experience.
Achievements
Apple WWDC25 Swift Student Challenge Winner
Global Winner
Won the prestigious Swift Student Challenge for chemAR, a spatial computing application that revolutionizes chemistry education through augmented reality.
ViewApple WWDC24 Swift Student Challenge Winner
Global Winner
Recognized by Apple for LimbRevive, an innovative rehabilitation app designed to assist amputees in their recovery journey using immersive technology.
ViewAtlassian Best PG Award Champion
University of Sydney
Awarded for presenting chemAR, where I integrated Gemini and RAG pipelines to enable intelligent, context-aware chemical data processing.
ViewAlgosQuest 24 Runner Up
iSource Innovations
Led team to 2nd place for building iSource, an AI-powered forum that automates query resolution using LLM agents, enhancing community engagement. (https://github.com/ineerajrajeev/iSource)
View3rd Place - Techfest IIT Bombay
AIQuest Hackathon
Secured 3rd place in Asia's largest science and technology festival for developing AIQuests, a robust AI-driven platform for automated quiz generation and adaptive learning. (https://github.com/Jeevanchoudhary9/AIQuests)
ViewAbout Me
I engineer intelligent systems that bridge the gap between human intent and machine execution. As a specialist in Agentic AI and Large Language Models, I focus on building autonomous workflows that solve complex, real-world problems.
My work spans the full stack of AI application development—from fine-tuning 7B+ parameter models and optimizing RAG pipelines to crafting immersive spatial experiences on Apple Vision Pro.
With a rigorous academic background from IIT Madras and the University of Sydney, combined with recognition as a two-time Apple WWDC Scholar, I bring a unique blend of research depth and product-focused engineering excellence. I am currently seeking opportunities to build the next generation of AI agents.
Experience
AI Research Intern
Indian Institute of Technology, RoparArchitected an autonomous evaluation agent using Multi-Agent Systems (MAS) and RAG. Deployed a fine-tuned Llama-3 model to grade subjective answers for 350+ students, achieving 98% correlation with human graders and reducing evaluation time by 60%.
Full Stack Engineer
Icon SteelEngineered a scalable CRM platform using microservices architecture. Implemented automated data pipelines processing 10k+ daily records and deployed on Azure Kubernetes Service (AKS) for high availability.
Education
University of Sydney
Master of Computer Science (Advanced Entry), AI Specialization
Feb 2025 - PresentIndian Institute of Technology, Madras
Bachelor of Science in Data Science & Applications
2020 - 2024University of Mumbai
Bachelor of Technology in Computer Engineering
2020 - 2024 Grade: 8.0/10Featured Projects
chemAR
An immersive Spatial Computing Agent for chemical education on Apple Vision Pro. Leverages hand-tracking and 3D object manipulation to visualize molecular structures in real-time. Won the WWDC 2025 Swift Student Challenge.
aiOS (Privacy-First)
A privacy-first Computer Use Agent (CUA) running entirely on-device (Mac MLX / Windows NPU). Architecture: Replaced vision-only methods with AX Tree parsing for 100x speedup and accuracy. Features Graph RAG (Neo4j) for system mapping, Qwen 3 Thinking VL for planning, and Function Gemma for tool use.
LimbRevive
An innovative rehabilitation platform designed to assist amputees in their recovery journey using immersive technology and spatial tracking. Winner of the WWDC 2024 Swift Student Challenge.
AIQuests
A robust AI-driven platform for automated quiz generation and adaptive learning, developed during Techfest IIT Bombay. Secured 3rd place among competitive teams.
Active Open Source Contributor