Skip to content

ArjunPakhan/arjun-intelligence-network

Repository files navigation

AIN — Arjun Intelligence Network

A portfolio built as an operating system, not a webpage.

Live: arjun-intelligence-network.vercel.app


Overview

AIN is a multi-layer intelligence system that functions as a personal portfolio for Arjun Pakhan — CS student at NMIMS Mukesh Patel School of Technology, specializing in Cybersecurity, Machine Learning, GPU Computing, and Distributed Systems.

The site is structured as an operational command network. Visitors don't browse a portfolio — they navigate a system.


Architecture — 8 Layers

Layer Name Purpose
0 Boot Sequence System initialization with particle hero
1 Command Center About + multi-agent system overview
2 Security Operations Cybersecurity projects + live threat map
3 Research Lab ML research + confusion matrix + feature explorer
4 Systems Thinking Intelligence briefs + reasoning tree
5 Memory Core Knowledge graph + research notes
5b Research Lab Hidden interactive ML demos
6 Terminal CLI contact interface

Keyboard shortcuts: 1–7 to navigate layers · T for timeline · R for recruiter mode · Esc to close overlays


Projects Featured

SENTINEL — Secure Distributed IDS

  • 3.5M CICFlowMeter network flow records
  • 10 attack classes · 99.7% detection accuracy
  • Stack: Java · Flask · Random Forest · XGBoost · Isolation Forest · AES-256 · FedAvg
  • IEEE paper submitted

CERBERUS — IoT Malware Detection

  • CIC-YNU-IoTMal 2026 dataset · 455K+ samples
  • 8 malware classes · ~88% XGBoost accuracy
  • 57% dimensionality reduction via hybrid feature selection
  • IEEE/Scopus submission in progress

UPI Gen Z Spending Behaviour

  • Neurofinance research paper (co-authored, 2024)
  • Key finding: 70% of Gen Z UPI transactions under ₹500
  • Peak fraud window: 7–10 PM
  • Guided by Dr. Sugam Shivhare, NMIMS

Tech Stack

Category Technology
Framework React 19
Build Vite 8
Styling Tailwind CSS 4
Animation Framer Motion 12
3D Three.js + React Three Fiber
Data Viz D3.js 7
Deployment Vercel

Content Architecture

All dynamic content lives in public/data/

public/data/ ├── certifications.json ← add new certs here ├── linkedin-posts.json ← add new posts here └── build-log.json ← weekly build updates no code changes needed to update portfolio content:

To update: edit JSON on GitHub → Vercel auto-deploys in ~30 seconds.


Local Development

npm install
npm run dev

Production build:

npm run build

Features

  • Live network threat visualizer with attack simulation
  • Interactive confusion matrix (IoT malware research data)
  • Feature importance explorer with 3 hybrid subsets
  • GPU vs CPU parallel processing race visualizer
  • UPI behavioral bias quiz (based on neurofinance research)
  • AES-256 CBC encryption live demo
  • D3 force-directed knowledge graph with timeline scrubber
  • Recruiter mode (press R) — role-filtered resume view
  • Dynamic resume download per role profile
  • Hidden research lab layer

Connect


Built between IEEE research papers.

Releases

No releases published

Packages

 
 
 

Contributors