Aditya Kumar
Anupam

Building systems at the intersection of Artificial Intelligence and Software Development.

View Resume View Work

GitHub LinkedIn Email

Based in India / LPU

Batch of 2027

About

I am Aditya Kumar Anupam, an AI Engineer and Full-Stack Developer. I architect intelligent applications by bridging advanced machine learning—including LLMs and agentic workflows—with scalable, production-ready software systems.

University

LPU, B.Tech CSE

2023 — 2027

Summer Training

Cipher Schools

May — Jul 2025

Expertise

Languages

PythonC++JavaJavaScript

AI & Machine Learning

PyTorchDeep LearningRAG SystemsAgentic WorkflowsLangChainGenAI

Full-Stack Engineering

FastAPIDjangoNext.jsFlutter

DevOps & Tooling

DockerPostgreSQLCI/CD PipelinesLinuxGitCloudflareAWS

Selected Work

Featured Systems & Engineering

All Work

PatentIQ – AI Patent Evaluation

Devised an end-to-end patent evaluation workflow automating ID extraction, validation, plagiarism analysis, AI scoring, and report generation. Built with Next.js, TypeScript, Cloudflare Workers, D1, and R2.

TypeScriptNext.jsCloudflareD1

Watermark Detector

Orchestrated a stateless watermark analysis pipeline using YOLOv8, FastAPI, and OpenCV for real-time in-memory image processing. Minimalist Next.js interface with side-by-side comparison views.

YOLOv8FastAPINext.jsOpenCV

Dia Chat – Fine-Tuned LLM

Curated a high-quality NLP dataset to train a personality-driven conversational model. Fine-tuned Gemma-3 1B Instruct with PyTorch and Unsloth. Deployed on-device via WebLLM and WebGPU.

PyTorchWebGPUGemma-3Unsloth

CrabClip – Secure Clipboard

A secure, anonymous clipboard for sharing content instantly with 4-digit OTPs. Features automatic TTL-based deletion, SHA-256 hashing, and distributed rate limiting to protect against brute-force attacks.

Next.jsMongoDBSHA-256Security

All Work