Hello, my name is
Shivani Pandey
I am a
MERN Stack Developer who loves building clean, fast web apps. Let's create something great together!
MERN Stack Developer who loves building clean, fast web apps. Let's create something great together!
Passionate and versatile tech professional specializing in Frontend Development, Cloud, Security, and Generative AI. I have hands-on experience building scalable web applications using Next.js, React, Node.js, and MongoDB, along with a strong foundation in AWS fundamentals and cybersecurity concepts.
I have started learning and am actively building my skills in Generative AI and LLM-based applications—including prompt engineering, AI agents, conversational AI, persona design, RAG systems, vector search, and tool-based workflows using LangChain—and I'm constantly expanding this expertise.
Alongside my technical background, I bring 3 years of teaching experience, strong communication skills, and a continuous learning mindset. I enjoy building user-centric, performant applications and experimenting with modern UI libraries like Shadcn UI, Material UI, Magic UI, and design systems to create clean and scalable interfaces.
Frontend · Backend · Cloud · Generative AI · Clean UI
Full-Stack AI Chatbot Application built with React.js, Node.js, Express, MongoDB, powered by OpenAI / Gemini API. QuickGPT allows users to chat with AI, generate text & images, manage credits, and explore a community gallery — all in a sleek, responsive interface.
Key features: User authentication (sign up & log in), real-time AI chat, AI image generation, credit system, community gallery, dark/light theme toggle. Responsive design, deployed on Vercel.
Tech Stack: React.js, Tailwind CSS, React Router, Node.js, Express.js, MongoDB, OpenAI / Google Gemini API, ImageKit, Stripe, Vercel
Wanderlust is a full-stack travel booking application built using the MERN Stack. It allows users to browse, search, and filter various accommodations. The platform includes secure user authentication, authorization (only owners can edit/delete listings), review and rating systems, dynamic image uploading via Cloudinary, and a fully responsive UI. It follows the MVC architecture and real-time updates ensure a smooth user experience. The backend is connected with MongoDB Atlas and frontend and backend are fully deployed on Render.
Tech Stack: Node.js, Express.js, MongoDB, Mapbox, Bootstrap 5, Cloudinary, EJS, JWT Authentication, Passport.js, Joi, Connect-Flash
Priscripto is a feature-rich full-stack web application designed for seamless appointment management. It provides dedicated portals for Users, Doctors, and Admins. Users can register, login, book appointments with doctors, and manage their bookings. Doctors can set available time slots, view upcoming appointments, and manage their schedules. Admins have access to oversee user and doctor activities, verify doctors, and manage the entire system. The application ensures secure authentication using JWT tokens and offers a smooth and responsive user experience across all devices. Real-time updates and robust backend APIs power the system, making it efficient and reliable.The project follows the MVC (Model-View-Controller) architectural pattern for better code organization and scalability.
Tech Stack: MongoDB, Express.js, React.js, Node.js, Cloudinary, React Toastify, Tailwind CSS, Axios, JWT (JSON Web Tokens), Mongoose, Git & GitHub
The Kohl's website clone is a comprehensive project built using React and Chakra UI and JS , showcasing advanced frontend and backend capabilities. It features secure user authentication and dynamic routing for smooth navigation. The real-time password feedback enhances user experience during account creation. Users can easily add items to their cart and delete them, with changes reflected instantly thanks to seamless integration with Axios. The project simulates a backend using JSON Server, deployed on Render, and efficiently manages user data with LocalStorage.
Tech Stack: HTML5, CSS3, JavaScript, React, Chakra UI, JSON Server, Axios, LocalStorage
An AI-powered customer support chatbot built as a hiring assignment for Spur Commerce Pvt Ltd. The application provides intelligent, context-aware responses for customer queries related to orders, shipping, returns, refunds, and company information. Conversations are persisted and restored using session IDs, and older messages are summarized to control token usage and cost.
The backend validates user input, rewrites unclear queries, matches FAQs when possible, and otherwise sends context-rich prompts to an LLM. The frontend supports streaming-style replies, typing indicators, suggestion pills, and automatic history loading for a smooth chat experience.
Tech Stack: Node.js, TypeScript, Express.js, React.js, Vite, SQLite, OpenAI API, REST APIs, LangChain Concepts, Render (Backend), Vercel (Frontend)