About Me
My Experience
Akamai Technologies
Developed a Python-based tool to identify and fix Split Brain issues in Akamai-MROM application's distributed database clusters.
Fond Solutions
Developed a high-rated smart pet collar mobile app using Flutter, improving backend response time by 40% and publishing on Google Play and Apple App stores
Rolling Right
Pioneered the development of a MERN stack E-Commerce web app with Shopify API, boosting sales by 15% and overseeing creative and coding aspects for production deployment.
My Skills
My Projects
Suicidal Intention Detection on Reddit
Developed a high-performance NLP model to detect suicidal intentions in Reddit posts, achieving 91.5% accuracy using LSTM networks and GloVe embeddings. The end-to-end machine learning pipeline processed 110,000 posts from various subreddits, implementing advanced text preprocessing techniques.
Sharify
Led a team of five in developing a full-stack web application for a Software Engineering course, creating a social media networking site where users can share the music they are listening to via Spotify links. Practiced Agile methodologies and utilized project management tools like JIRA. The project involved implementing user authentication, real-time chat system, and integration with the Spotify API to fetch and display song information.
Clustering of Corneal Epithelium Images
Completed a project on clustering CC3D simulated corneal epithelium images to classify them based on cell layer thickness. By utilizing UNet for feature learning and segmentation, followed by KMeans clustering, successfully categorized the images and enhanced image segmentation to improve clustering accuracy.
Best Venue Suggestor
Developed a real-time app that suggests optimal venues by considering distance, traffic, and user preferences. The app uses an NLP model with 84% accuracy to determine individual/group personalities based on the Myers-Briggs Test, and a model that optimizes central locations using real-time Google API data within three iterations. Both models are integrated using a cosine similarity test to find venues with match rates from 0 to 100%.
OMR Answer Detection
Developed a system for detecting and analyzing question boxes in scanned answer sheets. Using image processing techniques, the system accurately identified question columns, estimated box sizes, and detected marked answers. The implementation included handwritten character detection and highlighted marked responses for improved visualization. The project achieved 100% accuracy on test images, demonstrating effective computer vision and problem-solving skills
ShuffleTruffle - Classification of Path Shuffled Images
Implemented a comprehensive study on various deep learning models for image classification using the CIFAR-10 dataset. The project explored architectures like ResNetX, VGG16, Vision Transformers, and DenseNet121, analyzing their performance and characteristics. Additionally, novel approaches such as D-shuffletruffle and N-shuffletruffle were investigated to assess model robustness against image patch shuffling
Y Combinator Database
Developed a comprehensive database system to analyze Y Combinator-funded startups, processing data on over 4,000 companies. Using SQL and advanced data modeling techniques, created a normalized database structure that enabled in-depth analysis of startup trends, founder backgrounds, and regional ecosystems. This project provided actionable insights for optimizing investment strategies, identifying high-potential startups, and understanding key factors contributing to startup success in the tech industry.
An Exploratory Study on Kaggle Competition Trend
As part of a team, I analyzed the Meta Kaggle dataset, which provides a comprehensive overview of data analytics, modeling, and visualization challenges. Our report explored trends in Kaggle competitions from 2010 to 2024, examining topics, programming languages, submissions, evaluation metrics, and factors influencing team performance.
My Leadership Activities
Lead Teaching Assistant
Assisted in teaching I365 and I211, courses focused on full stack development using Javascript and Flask, and provided students with an introduction to Git and other essential developer tools. Facilitated learning of web application development, routing, templates, and version control.
PESU I/O Peer Instructor
As a Peer Instructor for the Flutter Development Course at PESU I/O, I conducted weekly instructional sessions, covering key topics such as Dart programming, widget creation, and state management. Additionally, I graded course assignments and projects, providing timely feedback, troubleshooting code issues, and offering continuous student support to enhance their learning experience.
Google Developer Student Club Core Member
As a core member of the Google Developers Student Club, I organized and conducted hackathons, fostering a collaborative and innovative environment. Additionally, I led introductory sessions on popular frameworks like Flutter and React, equipping students with essential skills and knowledge to excel in modern web and mobile development.