About Me
I'm a 3rd year Computer Science undergrad at People's Education Society (PES) University, Bengaluru.
My interest lies in Machine Learning, UI/UX, Systems programming. I look forward to working on solutions for real world problems !
Hello, I'm Nidhi !
Welcome to my little corner of the digital world :)
About Me ! Explore My WorkMy Projects
TrasHTTP
A lightweight, multithreaded web server written in C, designed to handle multiple client requests simultaneously, using configurable scheduling. Has support for HTTP and HTTPS. The user can choose from 3 different scheduling algorithms - Round Robin, least connections and random. Ensures static file serving with appropriate MIME type detection.
Water Meter in React Native
Developed a simple water meter app for an apartments association to track and manage water level information. Utilized React Native for cross platform compatibility and leveraged Android SDK within Android studio. Designed intuitive UI and UX aligned with Material Design principles.
Cloud based attendance system using RFID
Integrated various sensors (RFID readers, LCD, keypad) and an ESP32-WROOM microcontroller along with a custom socket program that leveraged the Wi-Fi capabilities of the ESP32 for real-time data transmission of attendance details and utilized an AWS EC2 instance for secure data storage and management.
Campus Skill Connect
Designed RESTful APIs for user authentication, skill/project management, and collaborations, with a Flask backend and MariaDB database. This skill exchange platform was built with React, Node.js, and SQL. It allows users to manage and collaborate on skills and projects of their choice, built for collaboration between peers in colleges who are new to coding.
AWS lambda Clone
Serverless function execution platform, re-creation of AWS Lambda's core idea. User can upload code (Python / JavaScript), execute that code in isolated environments (like Docker containers or MicroVMs) and track execution results, performance, and failures while monitoring everything live with dashboards.
This Portfolio
My favourite project of them all. Used simple HTML, CSS and Javascript along with Supabase. Used Supabase mainly to function as a CDN, to render the images and videos in the "Hobbies" section faster. The cute cats were my first attempt at pixel art using GIMP (GNU Image Manipulation - an open source raster graphics editor). Also, hosted it on vercel.
Work Experience
Machine Learning Research Intern
PESU Research Centre - ISFCRJune 2024 - August 2024
Conducted research on plant disease detection using image processing techniques and advanced machine learning algorithms like autoencoders, self-organizable maps, LightGBM and CatBoost. Utilized Python, TensorFlow, Keras, and Scikit-learn, along with data visualization libraries like Matplotlib. Used Roboflow to generate annotations for the tea-leaf dataset. Research published in the 5th International Conference on Advances in Electrical, Computing, Communication and Sustainable Technologies (ICAECT) - 2025
Systems and Software Engineering Intern
Drivool Technologies Pvt. Ltd.January 2023 - Present
Industrial IoT Control System with AWS-Hosted MQTT Architecture
Tools/Software/Protocols: MQTT Protocol, ESP32, ArduinoIDE, Wireless Technologies (Wi-Fi, BLE), AWS EC2, Node.jsProject PowerTap - Cloud connected Smart Electric Meter
Tools/Software/Protocols: UART, STM microcontroller, ESP32, ArduinoIDE, Wireless Technologies (Wi-Fi, BLE)Tested firmware for real time electricity monitoring. Focused on firmware-level debugging, system-level integration, and functional testing to ensure reliable cloud connectivity.
Exploration and hosting of microservices on an AWS EC2 instance
Tools/Software: AWS EC2, socket programming, MongoDBHosted microservices on an AWS EC2 instance and exposed the API. Developed and utilized multiple APIs for storing, retrieving, and deleting historical vehicle movement traces in MongoDB, hosted on an AWS EC2 instance, enhancing data management for GPS-based IoT solutions. Additionally, wrote cronjobs for deleting old historical data.
Research Work
Comparative Analysis of Machine Learning Algorithms for Binary Classification of Tea Leaf Diseases
Published in the 5th International Conference on Advances in Electrical, Computing, Communication and Sustainable Technologies (ICAECT) - 2025
Quantifying Bias in Vision Language Models
Existing multimodal architectures often exhibit a dynamic, question-dependent bias towards one modality over the other. This modality bias can lead to suboptimal performance and an imbalanced utilization of available information. There remains a need for a mechanism to distribute attention fairly to the modalities.
My Hobbies
Fine Arts
Click here to see some of my artwork over the years :)
Music
I am classically trained in Hindustani music! Click here to see some of my music covers.
Photography
Its mostly just flowers lol
Gaming
Indie games, farming sims, and puzzle games are my favorites. The cute cats on this webpage are my first attempt at Pixel art using GIMP !
Blog & Thoughts
Coming Soon! ✨
I'm planning to start a blog where I'll share my thoughts on technology, coding adventures, research insights, and maybe some life updates. Stay tuned for interesting content about the intersection of technology and creativity!
Topics I'm excited to write about: Web development tips, AI ethics, sustainable tech, and my journey as a CS student! 🌸
💌 Let's Connect!
I'd love to hear from you! Whether it's about collaboration, opportunities, or just to say hi 🌸
