Developed an online auction system which allowed sellers to post auctions for an item and potential buyers to bid for the product. Provided additional features such auto-bidding, view bid history, website statistics, product recommendations, etc. Designed a robust system to handle multiple concurrent requests and also provided bid alerts, new bids updates and item availability.
Engineered an advanced Online Account Management System in C++, integrating financial transaction capabilities with a QT Creator GUI for enhanced user experience. Features include real-time balance tracking, stock portfolio management, and secure transaction history. Utilized Singleton and Strategy patterns for efficient operation, and implemented synchronization across accounts. Tech Stack: C++, Ggplot, QT Creator
Led the development of a Document Image Classification project by integrating visual and textual features using machine learning models like VGG-16 and YOLO, and text analysis techniques such as TF-IDF and TextRank. Utilized OCR technologies along with Python libraries including Numpy, Pandas, Scikit-learn, Keras, Tensorflow, and PIL, significantly enhancing classification accuracy and efficiency.
Developed a movie recommendation system utilizing natural language processing techniques and cosine similarity to suggest movies to users based on their input. The system employs a dataset containing movie details such as keywords, cast, genres, and director, and utilizes these features to calculate similarity scores between movies.The recommendation system enhances user experience by suggesting relevant movies based on their preferences, thereby facilitating discovery and enjoyment of new content.
Users can search for specific questions Conceptual/Coding using OpenAI APIs. The interface lets users retrieve source code in the R language for customized dataset and provides R Code Editor to run the retrieved code. Experimented with ChatGpt 3.5 Turbo API.
Designed and developed a write-once file system with session based semantics from scratch(in C). Made efficient utilization of available memory to build disk-like structures. Implemented mount, unmount, read, write, open and close function calls to disk. Reduced external fragmentation. Implemented secured policy to detect broken disk structure and illegal file operations.
An android application and bidding platform that allows for efficient farmer-vendor interaction. Wrote efficient database queries - (SQLite). Developed the entire backend logic. Integrated Google maps API to calculate the delivery cost based on farmers location.
Designed and developed a custom memory allocator similar to malloc in C. Developed a mechanism to allocate, update, manipulate memory. Implemented segmented strategy and first-fit selection algorithm to represent memory. Tested the allocator with respect to memory saturation and time overhead.(Unix, C)
Worked on GET, PATCH and POST API’s for e-commerce websites using plain old Javascript and Express JS. Wrote the logic for adding an item to the cart by processing a PATCH request call. Established proper error handling mechanism and also tested the API’s (Node Js, Express JS, React, HTML5, CSS3).
Developed a real-time distributed collaborative editor using redis pub-sub messaging system and Javascript. Configured an HAProxy load balancer that would distribute the traffic across multiple servers. Performed tests for eventual consistency and conflict free writes/edits. (Python, Javascript, Express, HAProxy, Redis)
Implemented ML algorithms for diabetes prediction. Performed Data pre-processing, data visualization and data analysis on input data. Tuned various hyperparameters to optimal scale in order to maximize accuracy. Used multiple ML Algorithms and compared their accuracies. Algorithms used :- SVM, Logistic Regression, Naive Bayes.
Developed an innovative User Level Thread library in C++ to explore the efficiency and flexibility of lightweight threading. This project entailed designing and implementing various methods and algorithms to manage threads independently of the OS kernel. The library offers customizable scheduling and synchronization mechanisms, enabling it to operate on multiple operating systems without reliance on kernel-level support