Purvansh Sahu — ML Systems Portfolio

Personal portfolio focused on applied ML systems engineering. Built with React.

Live: purvanshsahu.vercel.app

What's in it

• Engineering Thesis — Positioning statement on building trustable ML systems
• Experience — ML research at IIT Madras (with ICAR-NRCB)
• Projects — Four production-oriented ML systems with technical depth:
  • PRGuard AI — Multi-agent PR review with confidence arbitration
  • Document Intelligence Engine — LayoutLMv3 + deterministic post-processing
  • GitHub Codebase Intelligence — Graph-aware RAG for code understanding
  • AuditLend — Credit decision engine with audit trail
• System Design — Architecture deep-dives with pipeline flows, failure paths, decision logic, and engineering tradeoffs
• Proficiencies — ML/Modeling, Systems/Backend, Tools/Infra

Tech Stack

• React.js
• GSAP + ScrollTrigger (scroll-linked animations)
• Lucide React (icons)
• Custom shader background (WebGL)
• Vanilla CSS

Run locally

git clone https://github.com/purvanshh/Portfolio-Machine-Learning.git
cd Portfolio-Machine-Learning
npm install
npm start

Opens at http://localhost:3000.

Structure

src/
├── components/
│   ├── Hero.js
│   ├── EngineeringThesis.js
│   ├── Experience.js
│   ├── Projects.js
│   ├── SystemDesign.js
│   ├── Proficiencies.js
│   ├── Education.js
│   ├── Contact.js
│   ├── Header.js
│   ├── Footer.js
│   ├── ShaderAnimation.js
│   └── ui/
│       └── HamburgerMenuOverlay.js
├── App.js
├── App.css
├── index.css
└── index.js

Contact

• GitHub: @purvanshh
• LinkedIn: Purvansh Sahu
• Email: purvanshhsahu@gmail.com

License

MIT