lsm-trees May 2025
Ground up implementation of LSM Trees and benchmarking against normal B-Trees
This project benchmarks the performance of B-Trees and Log-Structured Merge (LSM) Trees, both crucial for database indexing. B-Trees excel in in-memory operations, providing fast insertion, deletion, and search capabilities, making them ideal for database and file system indexing. LSM Trees, optimized for write-heavy workloads, periodically flush data from memory to disk as SSTables, handling large datasets effectively despite slower read times due to multiple SSTables. The benchmarking tests, implemented in C++, measure insertion and search times for both structures, highlighting their performance differences.
The project’s structure includes header and implementation files for B-Trees, LSM Trees, and SSTables, along with a main benchmarking file. Compilation and execution are simple, with commands provided for building, cleaning, and running the project. Results show B-Trees’ efficiency in memory-based operations, while LSM Trees are better for large datasets with frequent writes. Future improvements include optimizing LSM Tree performance through compaction, merging, and enhanced I/O operations.
Links: GitHub
dockermake November 2023
Add a makefile for your docker
In the software development world, Docker commands are essential but often cumbersome and unintuitive, particularly for new users. This complexity prompted the creation of “dockermake”, an npm package that simplifies Docker management using Makefiles.
By converting lengthy Docker commands into simpler “make” commands, dockermake streamlines the workflow, reducing the chances of human error. The package, inspired by a colleague’s Makefile, gained significant attention on Reddit, highlighting a common need for more efficient Docker command handling. Currently in beta, dockermake is open for contributions and aims to make Docker usage more accessible and user-friendly.
teachify November 2023
Fully administered AI tool for professors and students
Teachify is an AI-powered tool designed to enhance the educational experience for both students and professors. Developed with the expertise of Prof. Judi, Teachify offers seamless integration and administration of educational tasks. The main developers behind this project are Usman Zaheer, Basil Yusuf, Pranay Harpalani, and Faiz Mustansar.
Built using modern technologies like TypeScript, Next.js, PostgreSQL, and Supabase, Teachify aims to streamline daily academic activities through intelligent automation. The project is publicly available for collaboration and contributions.
Links: GitHub
fiable May 2023
Incentivized Biking through rewards
Fiable is an innovative app designed to reduce air pollution by incentivizing people to bike for shorter distances, offering monetary rewards for choosing this eco-friendly mode of transportation. Inspired by a Y Combinator request for solutions in transportation, the idea for Fiable emerged from brainstorming sessions and user feedback, aiming to make personal transportation more sustainable and enjoyable.
Development involved initial brainstorming and low-fidelity wireframe creation on Figma, followed by coding the app using React Native. The team faced challenges such as ensuring smooth navigation and real-time location tracking, but they successfully integrated Google Cloud Platform APIs and developed a user-friendly interface.
Achievements include a minimalistic design that enhances the user experience, seamless integration of location services, and creating an app that promotes environmental consciousness. Future plans for Fiable involve improving state management with Redux, enhancing polyline functionalities, and making design and security improvements. The project, developed by Basil Yusuf, Naol Denko, and Rashi Mathur, won the People’s Choice and Best Use of Google Cloud awards at GryphHacks 2022.
beachify June 2022
Centralized beach activity planning
In the realm of software development, planning and enjoying summer trips can be daunting, which inspired the creation of “Beachify,” an all-in-one summer hack. This project leverages blockchain technology and web3 authentication to streamline beach and park planning, making it easier for users to enjoy their outings.
Beachify simplifies the process by fetching beach information using Yelp’s Fusion API, integrating Google Maps for location checks, and rewarding users for their plans via the Hedera Test Network. Built with the MERN stack, Beachify offers a seamless user experience with a clean UI and decentralized authentication. Despite the challenges faced, including integrating Hedera and Google APIs, the project has received accolades for its innovative approach. Beachify is open for future enhancements to further integrate blockchain transactions and refine its user interface.
blindspot June 2022
Live web scraper to get unfiltered indian news
In the fast-paced world of news and current events, “Blindspot” offers a comprehensive and streamlined platform to keep users updated with the latest headlines. By aggregating news from reliable sources like NDTV, Blindspot ensures users have access to important updates, from political developments and social issues to celebrity news and health alerts. With an intuitive interface, it allows users to quickly navigate through different sections such as Left, Centre-Left, Centre, Centre-Right, and Right, making it easy to find news that aligns with their interests.
Blindspot also emphasizes user engagement and accessibility. It features a clean design powered by Tailwind CSS and offers seamless integration with social media platforms like Twitter and GitHub, allowing users to follow updates and contribute to discussions. The platform prioritizes transparency and user privacy, adhering to strict Terms of Service and Privacy Policies. As a one-stop news hub, Blindspot aims to keep its audience well-informed and connected to the world’s most pressing stories.