Final-year B.Tech (AI/ML) student who ships measurable, production-oriented systems — a Hybrid RAG chatbot scoring 0.92+ across all RAGAS metrics, a real-time bot-detection API at <5ms latency, an identity-reconciliation backend with full test coverage, and a from-scratch Kalman-filter multi-object tracker — spanning NLP/RAG, computer vision, cybersecurity AI, and full-stack development. Seeking an AI/ML or Software Engineering internship / full-time role.

Languages: Python, C++, C, JavaScript, HTML/CSS

AI / ML: PyTorch, TensorFlow, Keras, Scikit-Learn, XGBoost, LangChain, LangGraph, RAG, OpenCV, YOLOv8, BERT, LLMs, Agentic AI, FAISS, BM25, SHAP, RAGAS

Web / Backend: React, Next.js, Node.js, Express, FastAPI, Flask, REST APIs, Streamlit, MongoDB, PostgreSQL, Supabase, ChromaDB

Tools & Infra: Docker, Kubernetes, Git, GitHub, GitHub Actions, Linux, Vercel, Jest

AI / ML Intern · CodeAlpha (Jan 2025 – Feb 2025)

• Built a disease prediction model using patient medical data and classification algorithms (Scikit-Learn) to estimate disease likelihood from health indicators.
• Built a credit score prediction model applying supervised learning on financial / credit history data to classify creditworthiness.

Python LangChain FAISS / ChromaDB HuggingFace FastAPI React RAGAS

• Developed an end-to-end Retrieval-Augmented Generation pipeline: PDF ingestion → chunking → Hybrid FAISS + BM25 vector store → LangChain RetrievalQA, evaluated using RAGAS (0.92+ faithfulness, accuracy, and relevance).
• Shipped a multi-turn React chat interface over a FastAPI backend for interactive, source-grounded document Q&A.

Node.js Express Prisma SQLite Jest

• Built a backend service that reconciles customer identities by linking contact records (email / phone) across orders, even when different combinations are used per transaction.
• Delivered production-ready code with full Jest test coverage (13 / 13 passing), a submission-ready README, and a demo script.

Python XGBoost Isolation Forest SHAP Flask REST API

• Built a real-time ensemble ML system that fingerprints malicious HTTP traffic using XGBoost + Isolation Forest with SHAP explainability.
• Achieved <5ms inference latency in production — deployed as a Flask REST API with full explainability reports per request.

Python NumPy SciPy (Hungarian Algorithm) OpenCV YOLOv8 Pytest

• Implemented a Kalman filter and SORT-style multi-object tracker from first principles in NumPy — state prediction, Joseph-form covariance update, and Hungarian-algorithm data association — validated with 20+ unit tests.
• Benchmarked against a One Euro Filter baseline on synthetic noisy trajectories, cutting RMSE by up to 31% and position jitter by up to 86% under high-noise conditions while maintaining zero prediction lag through simulated occlusion.

B.Tech — Artificial Intelligence & Machine Learning, NIMS University, Jaipur Coursework: Deep Learning, NLP, Computer Vision, Machine Learning, AI, DSA, DBMS, OS, Computer Networks & Security

• Software Engineering Job Simulation — JPMorgan Chase & Co. / Forage — Jun 2026
• Introduction to Model Context Protocol — Anthropic (ID: 5utqn2pfrk7n) — Jun 2026
• Getting Started with AI on Jetson Nano — NVIDIA Deep Learning Institute — Dec 2025
• Introduction to Generative AI Studio — Google Cloud / Simplilearn — Dec 2025

• 🏆 QuizOff 2026 Finalist — India's Biggest AI Quiz (CampusCrew, hosted on Unstop), competing among 5,25,000+ students from 48,500+ institutions across 35+ countries — Jul 2026
• 🚀 Shipped 4+ end-to-end ML, computer vision, and backend/LLM projects within 6 months, from data pipeline to deployment
• 🤝 Active open-source contributor under the Agentic-IQ GitHub organization since April 2026
• 🏅 Certified in Anthropic's Model Context Protocol (MCP), NVIDIA Deep Learning Institute, and Google Cloud Generative AI