Hey, I'm Neel. I'm a software engineer and Georgia Tech CS student focused on applied AI and iOS. I'm currently building ML and graph analytics systems at GTRI's CIPHER Lab and shipping production-level SwiftUI apps with Core ML.
I build practical AI and iOS systems that translate research into products people can use every day.
Incoming Software Engineer Intern at GTRI CIPHER Lab focused on ML systems for cybersecurity and intelligence analysis.
Built and deployed AI-powered products including a multimodal image search engine over 10,000+ assets.
Led iOS product development with SwiftUI and Core ML, combining research-grade models with production UX.
My engineering philosophy stems from the idea that algorithms are only as good as the interfaces that present them. It's why I've dedicated myself to understanding the full pipeline—from crafting underlying embedding models with PyTorch, down to building reactive views with SwiftUI. Currently a third-year undergrad at the Georgia Institute of Technology, my focus is turning academic AI theory into tangible products.
When I'm not configuring a FAISS vector database or debating the latest Apple frameworks inside the GT iOS Club, you will likely find me staying active at the gym, tinkering with new Apple Silicon hardware capabilities, or finding ways to inject machine learning into everyday inefficiencies.
"Good code solves the math. Great engineering solves the user's problem."
Bachelor of Science in Computer Science
Atlanta, GA · GPA: 4.0 · Expected May 2027
Select a project below to read an in-depth case study of the architecture and implementation.
A full-stack live transit tracker: a 24/7 FastAPI/SQLite backend polling MARTA's GTFS-Realtime feeds (575K+ arrival observations across 78 routes and 6,200+ stops), feeding a native SwiftUI app tracking ~200 live vehicles at a 15-second refresh.
An extensible prompt-injection evaluation framework testing local and frontier LLMs across 46 payloads and 41 techniques, with a dual rule-based + LLM-as-judge scoring pipeline that quantified up to a 47% robustness gap on local 8B models.
A visual search engine indexing over 10,000 images using CLIP embeddings and FAISS, enabling retrieval from natural language prompts in under 2 seconds.
An elegant SwiftUI macOS/iOS application leveraging on-device Core ML to perform NLP semantic searching and dynamically generate knowledge graphs from user notes.
A computer vision pipeline using MediaPipe to accurately predict and interpret continuous American Sign Language streams into real-time synthesized speech.
PyTorch, TensorFlow, scikit-learn, FAISS
pandas, NumPy
Swift, SwiftUI, Core ML, MVVM
SQL, MySQL, MongoDB, Firebase, Node.js
Python, Java, C, C++, JavaScript, Next.js, React
Git, GitHub, CI/CD, Jenkins, GitHub Actions, Docker, CMake
AWS, Azure
A comprehensive overview of my education, experience, projects, and technical skills—available to view and download.