Aspiring Software Engineer & ML Engineer studying Computer Science + Applied Mathematics at Stony Brook University.
I build systems across computer vision, LLM/RAG pipelines, and full-stack applications, with experience in multi-view geometry, segmentation pipelines, agentic AI systems, and ML-driven applications.
- Machine Learning Intern @ IAMBIC — multi-view reconstruction, segmentation pipelines, 3D mesh generation, CNN+MLP modeling
- ML Engineer Intern @ 7-Eleven x Break Through Tech AI — LLM-powered RAG pipelines, LangChain, FAISS, multi-agent reasoning
- Break Through Tech AI Fellow — supervised/unsupervised ML, deep learning fundamentals, and applied ML workflows
- Interested in ML systems, computer vision, AI agents, and full-stack AI product engineering
- Languages: Python, Java, C, JavaScript, OCaml, SQL
- ML / AI: scikit-learn, NumPy, Pandas, CNNs, NLP, Gradient Boosting, K-means, RAG, LLM Integration
- Computer Vision: OpenCV, Multi-View Geometry, Mesh Processing
- Full Stack: React, Next.js, FastAPI, Tailwind, Zustand, Dexie.js
- Tools: LangChain, FAISS, VS Code
| Project | Description | Tech |
|---|---|---|
| Agentic-AI-for-Insights | End-to-end RAG + multi-agent pipeline for extracting insights from large enterprise documents; OCR → embeddings → FAISS search → multi-step reasoning; evaluated for retrieval precision (>80%) and clarity (>4/5). | LangChain, FAISS, Python, OCR, Embeddings, Multi-Agent Reasoning |
| u-mi-planner | React-based productivity and scheduling system with drag-and-drop groups/objects, 15-minute calendar scheduling, conflict detection, custom object pages, and offline persistence. | React, Zustand, Dexie.js, dnd-kit, TailwindCSS |
| ai-application-helper | Full-stack AI tool for resume/cover-letter improvement and tailoring; PDF/DOCX ingestion, text extraction, and LLM-based generation. | Next.js, FastAPI, Gemini API, PyMuPDF, python-docx |
- Stony Brook Computing Society (ACM Chapter): Workshops on ML, software engineering, UI/UX design, and collaborative project events
- Staying active: Rock climbing, basketball, and the occasional competitive ping-pong match
- Personal interests: ML systems research, CV pipelines, agentic LLM workflows, and building tools that improve productivity and learning
