|

Student at VIT, Vellore

Machine Learning | Web Development | App Development

GitHubLinkedInView Resume

About Me

I am a third-year undergraduate student at Vellore Institute of Technology with a strong passion for technology and innovation. My interests span across front-end development and machine learning.

Driven by curiosity and a growth mindset, I am always keen to explore new technologies—whether it's a modern framework, a programming language, or a tool that can enhance development workflows. Outside academics and projects, I enjoy gaming and continuously learning across a range of topics that pique my interest.

I am open to collaboration and discussions related to app development, machine learning, or any exciting idea that blends creativity with technology. Feel free to email me at shaneshaji10@gmail.com or reach out to me on my Linkedin!

GitHubLinkedInView resume

Skills

Projects

Fine Sentiment Analysis
Fine Sentiment Analysis
A full pipeline for multi-class sentiment analysis using neural networks and DistilBERT embeddings. Given a dataset of text reviews, it preprocesses the data, extracts embeddings using a pre-trained transformer, trains a custom neural network classifier, and evaluates its performance.
torchNLPJupyterhuggingface
Local PDF Chat RAG App
Local PDF Chat RAG App
This is a simple PDF question-answering app built with Langchain, on a local language model running through Ollama. You can upload any PDF file and ask questions about it and the app will pull out relevant information and give short, helpful answers.
streamlitLangChainOllamaRAG
HerWay
HerWay
Created a safety-focused MVP in a 2-day hackathon using React Native, allowing women to review and rate how safe different locations are. and rate how safe different locations are. The app used Firebase to store ratings and displayed heatmaps on native iOS and Android maps to visually represent safety levels.
React NativeExpoFirebase
Breakout Stock Screener
Breakout Stock Screener
A Python-based stock screener inspired by Mark Minervini's Volume Dry-Up (VDU) and Volatility Contraction Pattern (VCP) concepts. The program identifies stocks exhibiting significant volume compression, signaling potential imminent breakouts ideal for swing trading setups.
PythonAPIrequestsPandas
ClearSky
ClearSky
A React app providing real-time weather updates with a user-friendly interface. Features include light/dark mode toggle, smooth routing using React Router, and responsive design
ReactAPIGitHub Pages