het prajapati

About

Computer Science undergraduate with an 8.9/10 CGPA and hands-on experience building AI/ML, cybersecurity, mobile, and full-stack systems. Focused on applied machine learning, intelligent security systems, and scalable software engineering.

Experience

• Cyber Security Intern
  InTrainz
  Jun 2025 – Aug 2025

  Completed a paid Cyber Security internship with an integrated industry training phase.

Education

• B.Tech
  REVA University
  Computer Science & Engineering (IoT & Cyber Security incl. Blockchain) · 2028

Skills

Tools / apps / platforms

Languages

Projects

• APARTMENT MANAGEMENT OS

  In-development apartment operations platform for independent and mid-sized communities, covering resident management, maintenance, billing/finance, complaints, and communication, with future AI-driven operations planned.

• SMARTSHOP — Full-Stack Product Management System
  React, Vite, Tailwind CSS, Spring Boot, Spring Security, JWT, MySQL

  Admin dashboard with a Spring Boot backend implementing protected product CRUD functionality.

• PHISHING SITE DETECTION — ML-Powered Browser Security Extension
  Python, Machine Learning, Flask, JavaScript, REST API, Feature Engineering

  Browser extension backed by a Flask ML API; classifies URLs via engineered indicators, combining trusted-domain validation with ML probability thresholds to return phishing/safe predictions with confidence scores.

• ECO-GUARDIAN — AI Environmental Hazard Reporting System
  Vision AI, React, Three.js

  Developed an AI-agent workflow that accepts environmental hazard images, performs Vision AI detection, and uses a second agent for cross-validation before producing confidence-scored reports. Built a React + Three.js interface with 3D Earth visualization, incident tracking, analytics, agent-health monitoring, and automated routing of verified reports.

• PROJECT 117 — Sovereign Industrial Intelligence Platform
  RAG, FastAPI, React, Ollama, LanceDB, SQLite, Docker

  Engineered a refinery digital-twin platform spanning 58 equipment assets, 224 sensors, 60 process connections, and 18 process areas, with fault simulation, event streaming, recovery actions, verification, and audit trails. Built a multi-agent incident-response pipeline combining RAG, topology-aware validation, independent verification, human approvals, and persistent audit logging; validated with 368 passing backend tests and a successful production build.

• SENTINEL.GRID — Federated AI Cyber Defense Platform
  CNN-LSTM, CICIDS2017, Flower, FedAvg, Differential Privacy, FastAPI, React

  Built a real-time network intrusion detection system using a CNN-LSTM model trained on CICIDS2017, achieving 98.85% accuracy across DDoS, port scan, brute-force, web attack, and normal traffic classes. Implemented Flower-based federated learning across 3 edge nodes with FedAvg and differential privacy, alongside automated IP blocking, rate limiting, and honeypot responses.

Courses & certifications

• Data Fundamentals · IBM SkillsBuild · 2024

🏆 Achievements & awards

• Adobe University Hackathon Participant · 2026
• Stack Fusion Fest Participant · 2025

  24-Hour Hackathon

• Code Quest 2.0 Phase 1 Participant · 2025