const Prakul: Developer = {
name: "Prakul Dhiman",
location: "India 🇮🇳",
education: "B.Tech — Computer Science Engineering",
role: "Software Engineer (Full Stack / Backend)",
building: "FraudShield — real-time AI fraud detection platform",
learning: ["System Design", "AWS", "Microservices", "Applied ML"],
philosophy: "Ship fast. Refactor smart. Scale confidently.",
stack: {
languages: ["JavaScript", "Java", "Python", "SQL"],
backend: ["Node.js", "Express.js", "Spring Boot", "FastAPI"],
databases: ["MongoDB", "PostgreSQL", "Redis", "MySQL"],
ml: ["XGBoost", "SHAP", "Pandas", "scikit-learn"],
frontend: ["React", "Tailwind CSS"],
tools: ["Git", "Docker", "Postman", "Swagger"],
},
openTo: "Full Stack / Backend Engineering Roles",
contact: "prakuldhiman@gmail.com",
};- 🛡️ Building FraudShield — a real-time credit card fraud detection platform with explainable AI (XGBoost + SHAP)
- 🚀 Also shipped IntelliHire — an AI hiring platform with live collaborative coding interviews
- 🌱 Deep-diving into System Design, AWS & Microservices
- 💬 Ask me about Node.js · Spring Boot · FastAPI · MongoDB/PostgreSQL · Applied ML
- 🤝 Actively looking for backend / full-stack roles — let's talk!
📊 Skill Proficiency Breakdown — Click to expand
| Skill | Level | Proficiency |
|---|---|---|
| Node.js / Express.js | Advanced | ██████████████████░░ 90% |
| MongoDB | Advanced | █████████████████░░░ 85% |
| JavaScript / TypeScript | Advanced | ████████████████░░░░ 80% |
| Java / Spring Boot | Intermediate–Advanced | ███████████████░░░░░ 75% |
| Python / FastAPI | Intermediate | ██████████████░░░░░░ 70% |
| React | Intermediate | ██████████████░░░░░░ 70% |
| Applied ML (XGBoost / SHAP) | Intermediate | █████████████░░░░░░░ 65% |
| PostgreSQL / SQL | Intermediate | █████████████░░░░░░░ 65% |
| Redis | Intermediate | █████████████░░░░░░░ 65% |
| Docker | Learning | ████████░░░░░░░░░░░░ 40% |
| AWS / Cloud | Learning | ██████░░░░░░░░░░░░░░ 30% |
Click any project to expand details, architecture, and code snippets.
🧠 A production-shaped fraud detection system that flags fraudulent credit card transactions in real time — and explains why each prediction was made.
✨ Features
| Feature | Implementation |
|---|---|
| 🧠 Explainable AI | SHAP breaks down which features triggered each fraud flag |
| ⚡ Real-Time & Bulk Scoring | Single transaction or CSV batch upload |
| 📊 Analytics Dashboard | Historical fraud trends via Recharts |
| 🔐 Auth | JWT + bcrypt + OTP phone verification |
| 🎨 UI | Custom glassmorphism dark theme, fully responsive |
🏗️ Architecture
┌──────────────────────────────┐
│ React + Vite (Frontend) │
└──────────────┬───────────────┘
│ REST API (JWT)
┌──────────────▼───────────────┐
│ Node.js + Express (API) │──── MongoDB (Mongoose)
└──────────────┬───────────────┘
│ Internal REST call
┌──────────────▼───────────────┐
│ FastAPI ML Microservice │
│ StandardScaler → XGBoost │
│ → SHAP TreeExplainer │
└───────────────────────────────┘
🛠️ Tech Stack
- Frontend: React 18, Vite, Tailwind CSS, Recharts, Framer Motion
- Backend: Node.js, Express.js, MongoDB, JWT, Bcrypt
- ML Service: Python 3.10+, FastAPI, XGBoost, SHAP, Pandas, scikit-learn
🧠 An AI-driven recruitment platform with live, real-time technical interviews and automated candidate evaluation.
🔗 Frontend Demo · 🔗 Backend API
✨ Features
| Feature | Implementation |
|---|---|
| 💻 Collaborative Code Editor | Monaco Editor + CodeMirror |
| ⚙️ Sandboxed Code Execution | Judge0 API — Java, Python, JS, C++ |
| 🌐 Live Interview Rooms | Real-time sync via Socket.io |
| 📄 AI Resume Processing | Automated parsing via pdf-parse |
| 📊 Candidate Scoring | Heuristic score from resume + coding + communication |
| 🎨 UI | Tailwind CSS, Framer Motion, Three.js accents |
🏗️ Architecture
React (Vite) Client
│
Node.js + Express API
│
MongoDB (Mongoose) ──── Redis (cache/session)
│
Judge0 API (external code execution)
🛠️ Tech Stack
- Frontend: React (Vite), Tailwind CSS, Framer Motion, Three.js, Socket.io-client, Monaco Editor, Recharts
- Backend: Node.js, Express.js, MongoDB (Mongoose), Redis
- Security: JWT, Helmet, Express Rate Limit, CORS
🔐 A secure distributed file management and sharing system with cloud-ready storage infrastructure.
🛠️ Tech Stack
- Backend: Java, Spring Boot
- Frontend: React
- Data: PostgreSQL, Redis
- Infra: Docker, Docker Compose
Full README and feature breakdown coming soon — actively documenting this one.
🔗 A MERN job portal connecting students and recruiters — built to learn full-stack development end-to-end.
✨ Features
| Feature | Details |
|---|---|
| 👤 Student Profiles | Registration & profile management |
| 🔍 Job Search | Search and apply, track applications |
| 🏢 Recruiter Dashboard | Post and manage job listings |
| 🔐 Auth | Secure authentication system |
🛠️ Tech Stack
HTML, CSS, JavaScript, React, Node.js, MongoDB
2021 ──────────────────────────────────────────────────────────────── Now
│
├── 📚 Started B.Tech CSE
│ └─ C basics → OOP → Data Structures
│
├── 💡 Discovered Web Development
│ └─ HTML · CSS · JavaScript · PHP
│
├── ⚡ Levelled Up — Modern Backend Stack
│ └─ Node.js · Express.js · MongoDB · REST APIs
│
├── 🔴 Added Caching & Performance Layer
│ └─ Redis · JWT Auth · Rate Limiting · Swagger Docs
│
├── ☕ Expanded Into the JVM World
│ └─ Java · Spring Boot · PostgreSQL (Atlas)
│
├── 🧠 Went Deep on Applied ML
│ └─ Python · FastAPI · XGBoost · SHAP (FraudShield)
│
├── 🚀 Building Production-Grade Apps
│ └─ FraudShield · IntelliHire · Atlas
│
└── 🎯 Current Goal: Land Backend / Full Stack Role
└─ System Design · AWS · Open Source Contributions
🏗️ How do you approach system design?
I follow a structured process:
- Clarify requirements — functional & non-functional (latency, throughput, availability)
- Estimate scale — DAU, read/write ratio, storage needs
- Design data model — SQL vs NoSQL based on access patterns
- Define APIs — REST contracts with clear status codes
- Identify bottlenecks — caching (Redis), CDN, DB indexing, sharding
- Document trade-offs — CAP theorem, consistency vs availability
⚙️ What's your strongest technical area?
Full-stack systems that combine a conventional backend with a real ML or real-time component — for example, FraudShield's Node.js API talking to a separate FastAPI/XGBoost service, or IntelliHire's Socket.io-driven live interview rooms. I'm comfortable owning the API layer, the data model, and the service that talks to them.
🧹 How do you maintain code quality?
- Structure: clear separation between client, API, and any ML/service layer
- Error handling: centralized middleware, no scattered try-catch
- Validation: input validation at the route level — never trust raw input
- Documentation: a real README per project, with setup steps that actually work
- Security:
.env.examplefiles, never committing real secrets, JWT + bcrypt for auth
🌱 What are you learning right now?
2025 Learning Roadmap
───────────────────────────────────────────────────────
✅ Applied ML — XGBoost, SHAP, model explainability
✅ Spring Boot — building a distributed system (Atlas)
🔄 System Design — HLD/LLD for scale
🔄 AWS — EC2, S3, Lambda, RDS fundamentals
📌 Testing — Jest, Supertest, TDD approach
📌 Kubernetes — container orchestration basics
───────────────────────────────────────────────────────
I'm actively looking for Full Stack / Backend Engineering roles where I can ship real impact at scale.
