Hanshika Duhoon

I’m a B.Tech (Information Technology) student focused on building full-stack applications and AI systems that translate complex models into useful products. I enjoy working at the intersection of software engineering and Generative AI, especially Retrieval-Augmented Generation (RAG).

In projects, I built an AI-powered food redistribution platform using MERN, Tailwind CSS, and JWT for secure role-based dashboards, along with geolocation-based NGO recommendations. I also created an ATS Resume Builder and Analyzer with Next.js/React/TypeScript that parses PDFs/DOCX (PDF.js, Mammoth.js) and scores resumes against job descriptions.

My AI work includes a RAG research assistant that answers questions from uploaded PDFs using LangChain and ChromaDB, with a Streamlit interface for semantic search and real-time responses. I’ve also conducted explainable hate speech detection research with LLMs on the MetaHate dataset, including a prompt engineering framework and evaluation (Vicuna-13B reached 80.95% accuracy).

Across internships, I developed mobile apps with Flutter/Dart and built full-stack web applications integrating machine learning models and REST APIs. I bring strong DSA momentum (300+ LeetCode problems, 150+ on GeeksforGeeks) and a collaborative mindset, from hackathons to organizing events for 200+ participants.