Web platform developed as part of work at OSPO for QDR at Syracuse University. QDR is a dedicated archive for storing and sharing qualitative and mixed-method research data. It hosts variety of collections one of them being HEAL

Worked to develop a automated pipeline between to transfer study-level metadata from HEAL and QDR

Involved extension JSON structure validation and transformation followed by API calls to upload data to dataverse (backend used by QDR)

This is a tool I developed as part of development at OSPO for Syracuse University. It aims to help faculty and students find the most appropriate data storage option based on their specific needs using a simple survey method.

This tool is built on WordPress as the content management system, featuring a custom-written plugin that renders the user interface thoroughly tested for functionality, responsive design and accessibility testing using WAVE

Conducted faculty interviews and collaborated with departments across university to collect list of storage solutions.

Built a mobile point-of-sale app for independent bookstore owners to simplify sales and inventory management. Developed the Android app using Kotlin with features like barcode scanning, transaction history, and real-time inventory updates.

Created the backend using Spring Boot and connected it to an SQL database. The backend is deployed on AWS EC2 to ensure reliable and scalable access. Persistent local storage for cart items using RoomDB

Designed to eliminate the need for expensive hardware, offering a cost-effective and accessible retail solution.

Detect cataracts and classify severity (mild, normal, severe) to improve early detection accuracy.

Achieved 96% accuracy in cataract detection using combined SIFT-GLCM features with SVM, Random Forest, and Logistic Regression. SIFT-GLCM individually showed lower accuracies. Attained 97.66% validation accuracy for severity classification using deep convolutional neural network models (SqueezeNet, MobileNet, VGG16).

Results accessible via web and Flask-based user interface. Phase 1 implementation published in ICAIAA 2021 conference paper for binary cataract classification.

The primary objective was to design a system that significantly reduces manual efforts by automating various processes related to farming.

Established sensor network within the farm premises, powered by Raspberry Pi, to monitor critical physical parameters including smoke, temperature, humidity, and moisture. This data is meticulously analyzed on cloud platforms to construct a predictive model for accurate weather forecasting. Additionally, camera-based surveillance is implemented to monitor and detect any trespassing incidents on the farm.

The outcomes and insights derived from these monitoring systems are conveniently accessible through a Flask-based GUI

To create web application leveraging the open source API to provide users with a seamless search experience for movies and TV series.

Using React, a user-friendly interface was crafted, enabling users to easily search, filter, and access detailed information about various titles. Redux was integrated to manage the application's state efficiently, ensuring smooth data flow and responsiveness.

The resulting web application offers users a streamlined experience for browsing movies and TV series. With Redux handling the state management, the application maintains performance and states.

An integrated Chatbot on platforms such as WhatsApp and Telegram. This Chatbot serves as a comprehensive tool to address admission-related inquiries without necessitating a physical visit to the college.

System developed using Google Dialogflow with knowledge base collected with the help of Faculty members at undergraduate Institute.

This idea was further copyrighted by faculty at the Department of Information Technology so as to scale for the university in the coming years.