Rohit Shrivastava
SDE Intern @ Solution Cone & CDMM
Building resilient web systems & spatial tech.
Hi, I’m Rohit Shrivastava. SDE Intern at Solution Cone & CDMM VIT, and Founder of the Geospatial Club. I engineer scalable web backends, disaster-warning visualization applications (ShakeAlert), and spatial intelligence.
Buildify
Featured IDE
3+
SDE Internships
5
Certifications
1 Paper
Published Research
ShakeAlert • Seismic Early Warning System
Integrated real-time seismic sensor feeds for instantaneous disaster mitigation & alert visualization at CDMM, VIT Vellore.
Core Backend
PHP 8 & CodeIgniter
Cloud & Space
AWS & DRR Tech
Production Code & Engineering Leadership
4 core projects spanning browser Web IDE runtimes (Buildify), agentic healthcare & disease prediction (Arogya AI), real-time crowdsourced disaster telemetry (Earthquake Detector), and multi-format AI content analytics (Social Media Analyzer).
Case Studies
Case Study 01 — AI & Browser Systems
Buildify Editor: AI-Powered Browser Web IDE
Case Study 02 — Agentic Healthcare & ML Runtimes
Arogya AI — Agentic Disease Prediction with Ayurvedic Recommendations
Case Study 03 — Crowdsourced Telemetry & Disaster Response
Real-Time Earthquake Detector & Disaster Response Demo
Case Study 04 — AI & OCR Content Pipeline
Social Media Content Analyzer
Buildify Editor: AI-Powered Browser Web IDE
A blazing-fast, AI-integrated Web IDE built entirely in the browser using Next.js 15 App Router, WebContainers, Monaco Editor, xterm.js, and local LLMs via Ollama.
Traditional cloud IDEs require expensive server infrastructure, introduce latency for full-stack application execution, and lack zero-latency local AI autocomplete capabilities.
Architected an in-browser Web IDE leveraging WebContainers for client-side Node.js execution, custom Monaco Editor integrated with local Ollama LLMs for instant inline completion, interactive xterm.js terminal, OAuth authentication, and multi-stack project templates (React, Next.js, Express, Hono, Vue, Angular).
Technical Architecture
Engineering Result
"Empowers developers to edit, refactor, chat with AI assistants, and execute frontend/backend microservices directly inside the browser with zero cloud server overhead."
Runtime
Browser WebContainers
AI Completion Engine
Ollama Local LLMs
Official Live IDE
buildify-eta.vercel.app
Arogya AI — Agentic Disease Prediction with Ayurvedic Recommendations
A hybrid healthcare assistant that predicts likely diseases from user-reported symptoms using a Multinomial Naive Bayes model, validates and refines predictions with an agentic rule layer, and provides culturally relevant Ayurvedic recommendations and human-readable explanations via an LLM.
Traditional symptom checkers lack intelligent verification layers to prevent false positives, miss holistic/ayurvedic health insights, and fail to provide transparent, human-readable medical rationale behind disease predictions.
Architected a hybrid ML microservice utilizing a Multinomial Naive Bayes classifier paired with an agentic rule layer for hypothesis validation and locking, integrated an Ayurvedic knowledge base for tailored recommendations, and integrated LLM explanations for human-readable diagnostic reasoning.
Technical Architecture
Engineering Result
"Delivers rapid, multi-stage symptom analysis with hypothesis locking, automated validation rules, actionable Ayurvedic guidance, and transparent LLM-generated diagnostic reasoning."
Classifier Model
Multinomial Naive Bayes
Validation Engine
Agentic Rule Layer
Knowledge Base
Ayurvedic KB + LLM
Real-Time Earthquake Detector & Disaster Response Demo
A resilient web application for real-time earthquake detection and disaster response, leveraging crowdsourced mobile accelerometer sensors, real-time Firebase services, Snapchat-style disaster heatmaps on Google Maps, and offline satellite messaging capability.
Traditional disaster response platforms lack real-time crowd-verification mechanisms to validate seismic activity instantly, leading to false alarms, delayed emergency dispatch, and complete blackout during cellular network outages.
Architected a client-side sensor engine monitoring device accelerometers, integrated Firebase real-time data sync for crowdsourced shake validation, designed an interactive Google Maps portal featuring Snapchat-style disaster intensity heatmaps with profile markers, and implemented offline chat & satellite messaging resilience.
Technical Architecture
Engineering Result
"Empowers communities and first responders with high-accuracy, crowd-verified seismic detection alerts, dynamic disaster intensity visualization, and fault-tolerant emergency communication."
Detection Engine
Accelerometer Sensor API
Verification Model
Crowdsourced Validation
Visual Analytics
Google Maps Disaster Heatmap
Interactive AI Web IDE & Browser Terminal
Simulating Buildify Editor (Vibecode Editor) — an AI-powered browser IDE built with Next.js 15, WebContainers, Monaco Editor, and Ollama local LLMs.
// Press "Trigger AI Suggest" below to see autocomplete in action
🚀 WebContainers engine initialized in 14ms
📦 Loaded Next.js 15 App Router runtime
🤖 Connected to Ollama Local LLM (qwen2.5-coder)
⚡ Server listening on http://localhost:3000
Click controls to test local AI suggestions or open the full live web application.
Technical Radar & Capabilities
Technologies and platforms factually applied across my SDE internships, disaster management projects, and leadership experience.
AI, Web IDE & Browser Runtimes
Full-Stack & Server Systems
Databases & Cloud Architecture
Geospatial & Engineering Leadership
Engineering & Leadership Track
Software development internships, disaster management application engineering, and technical community leadership.
Chairperson & Founder • GEOSPATIAL CLUB, VIT Vellore
- Founded the university's premier Geospatial technical society dedicated to GIS and satellite data engineering.
- Led cross-functional teams, organized technical workshops, and promoted spatial software innovation.
- Organized student initiatives bridging computer science with remote sensing concepts.
SDE Intern • Solution Cone
- Designed, developed, and maintained dynamic web applications using PHP, MySQL, and CodeIgniter framework.
- Built production-ready CRUD modules and enforced strict MVC architecture for clean, maintainable codebases.
- Optimized backend queries and database operations for enhanced web application responsiveness.
SDE Intern & PHP Tutor • Varcity Services
- Built end-to-end web modules including authentication systems, user administration dashboards, and data features.
- Collaborated directly with client teams for business requirement gathering and technical specification.
- Mentored student developers in server-side engineering, Core PHP, MySQL, and secure software development methodologies.
SDE Intern — CDMM • Centre for Disaster Mitigation and Management (CDMM), VIT
- Worked at CDMM VIT developing web technologies and user interfaces for disaster mitigation and data visualization.
- Designed responsive UI components for real-time data visualizers.
- Implemented backend logic to process web data streams and generate notification modules.
Published Research & Empirical UX
Technology Acceptance and Behavioural Patterns of Educational App Use Among Indian Engineering Students: An Empirical Study
This empirical study investigates the behavioral adoption, usability drivers, and structural technology acceptance factors among Indian engineering undergraduates utilizing digital education platforms. Through quantitative survey models and statistical analysis, the paper identifies core UX determinants that drive long-term student engagement.
Certifications & Technical Badges
AWS Certified Solutions Architect
Amazon Web Services (AWS)
ISRO-Overview of Geographical Information System
Independent Software System
Software Engineering Job Simulation
Forage / Industry Simulation
Master Search Engine Optimization (SEO)
Advanced Web Marketing Academy
Strategic Thinking for Leaders
Leadership Institute
Social Media Content Analyzer
A production-quality full-stack application that extracts social media content from PDFs and images (PNG, JPG, JPEG) using PyMuPDF and Tesseract OCR with semantic block preservation, normalizes text, and analyzes engagement using an AI engine (Gemini API / Fallback).
Extracting social media content from heterogeneous PDF documents and image scans often destroys document hierarchy, introduces OCR noise, and fails to provide structured, actionable engagement insights.
Architected an async FastAPI and React/TypeScript pipeline using PyMuPDF coordinate block parsing, Pillow preprocessing + Tesseract OCR fallback, noise filtering, and a hybrid Gemini API / heuristic AI scoring engine for engagement, hook strength, and CTA optimization.
Technical Architecture
Engineering Result
"Delivers multi-format PDF & image text extraction with semantic block preservation, 0-100 score integrity (Hook, Readability, CTA, Emotion), structured document toggle UI, side-by-side rewritten copy comparison, and instant hashtag generation."
Extraction Engine
PyMuPDF & Tesseract OCR
AI Analytics Engine
Gemini API + Fallback
Supported Formats
PDF, PNG, JPG, JPEG (<=10MB)