kushal = {
    "name"        : "Kushal Choudhary",
    "degree"      : "B.Tech CSE (AI & Robotics) @ VIT Chennai",
    "graduation"  : 2026,
    "focus"       : ["Backend Systems", "Machine Learning", "Generative AI"],
    "languages"   : ["Python", "C++", "JavaScript", "SQL"],
    "currently"   : "Building production-ready AI-powered systems",
    "fun_fact"    : "I can debug at 2AM and still push clean commits ☕"
}

Languages

Frontend

Backend & Databases

AI/ML

Cloud & Tools

Python Graph Neural Networks XGBoost SHAP

A hybrid fraud detection engine that models transactions as temporal graphs to uncover complex fraud patterns. Reduced false positives by 25% through advanced feature engineering. Integrated SHAP for full model explainability.

Node.js MongoDB Webhooks

Real-time backend platform for e-commerce analytics. Built event-driven data pipelines for order/revenue tracking with webhook ingestion, retry logic, and scalable API design.

Python scikit-learn Flask Pandas

End-to-end ML pipeline predicting sleep disorders (Insomnia / Sleep Apnea) across 400+ patient records. Trained a Random Forest classifier with cross-validation and deployed via a Flask web app with an interactive UI.

Extern — Artificial Intelligence & Machine Learning · SmartBridge (Aug 2023 – Nov 2023)

Built QueryCraft, an AI-powered Text-to-SQL platform that improved SQL accuracy by 30% using prompt engineering and schema grounding. Designed full query generation, validation, and execution pipelines.