turning code, curiosity, and caffeine into creation
I'm a B.Tech student in Information Technology and Mathematics at the Cluster Innovation Centre, University of Delhi — where logic meets innovation (and occasionally, late-night debugging marathons). My work lies at the intersection of AI, accessibility, and quantum computing — from developing intelligent assistive technologies that empower visually and hearing-impaired learners, to exploring quantum algorithms that push the limits of classical computation. You could say I like making things super — whether it's super-position or super-useful.
Along the way, I've had the opportunity to work as a Student Research Intern at IIIT Delhi, where I explored the scalability and efficiency of quantum algorithms, and as a Student Intern at Hindu College, where I applied transformer-based deep learning models to solve complex scientific prediction problems. At the Cluster Innovation Centre, I'm currently designing AI-driven assistive solutions that integrate speech, audio, and Braille modalities to make learning more inclusive and adaptive for people with multiple impairments.
I also have a strong foundation in backend development and UI/UX design, though lately, my interests have gradually shifted towards post quantum cryptography and quantum safe encryptions.
Beyond technology, I love dancing to the rhythm of Odissi and playing the violin — because whether it's algorithms or art, I've always believed that true creativity lies in finding harmony between logic and expression.
Delhi, India
Here are some of the projects I have recently worked on
Currently developing an AI-powered waste classification system using deep learning to automatically categorize trash into 7 distinct types (cardboard, e-waste, glass, medical, metal, paper, and plastic). The project leverages transfer learning with a pre-trained ResNet-34 architecture, implementing a two-phase training strategy where the model first learns with a frozen backbone before fine-tuning the entire network.The model is being developed for potential deployment in automated recycling facilities to streamline waste sorting processes and improve recycling efficiency.
Led the design and development of a machine learning–based Fraudulent Job Posting Detection System, created DFD, ER, and UML diagrams, and performed Function Point and COCOMO estimations to ensure a scalable, cost-efficient architecture. Currently leveraging web scraping and ML models to analyze job postings in real time, detect fraud through red flags (vague descriptions, unrealistic salaries, missing websites, personal info requests), and provide users with tailored updates on recent fraudulent listings in their fields of interest.
Implemented a quantum image encryption algorithm using Qiskit to enhance security through quantum superposition and entanglement. The encryption method leveraged the NEQR model, achieving 100% accuracy without any data loss, and significantly increased data protection complexity. By utilizing advanced quantum circuit techniques, the approach effectively reduced potential decryption vulnerabilities. The project involved technologies such as Qiskit, Python, Quantum Circuits, and NumPy.
A back-end based project using Django framework developed for an efficient project management app designed for team collaboration, task tracking, and milestone management. Features include user authentication, role-based access, and a responsive dashboard for efficient tracking of projects and tasks. One will be able to keep a track of the project list and assign tasks under the project in order to ensure which are the tasks that are new, completed, in progress, rejected, fixed and passed.
Developed a to-do list site which enables the user to make schedule for the day along with time and task and write their individual top priorities. It also allows them to set challenges for themselves and write down their mood for the day. It helps them in organizing tasks that are to be done soon and the day after offering users to set the theme according to their preference.
Developed a backend system for a restaurant named LittleLemon containing Menu and Booking API connected to a MySQL database. The Menu API helps to view, add and delete different items and Booking helps to register the number of guests along with the time and date they would be visiting the restaurant.
Developed a Flask-based URL shortener application that generates and manages shortened URLs efficiently. The system extracts keywords from URLs to create meaningful short codes, with a fallback to random codes when necessary. A database ensures persistence by storing both original and shortened URLs, while a redirection mechanism seamlessly routes users to the original link upon access.