Projects

Things I’ve built

A collection of side projects, experiments, and shipped products across AI/ML, web, and mobile.

All Web Natural Language Processing Artificial Intelligence Deep Learning Mobile Cybersecurity IoT Computer Vision Backend Games Research Quantum Data DevOps Robotics

Apr 2025

E-Paper Photo Frame

E-Paper Photo Frame is a smart digital photo frame built using an e-paper (e-ink) display and an ESP32 microcontroller. It connects to the internet and syncs with a web portal where users can manage the images displayed. The system supports optimized photo rendering, slideshows, and a clock mode for displaying time.

ESP32E-Ink DisplayIoTNext.js

Mar 2025

Gen AI-Orchestrator for Email and Document Triage/Routing

This project automates the classification and processing of high-volume email requests, especially in domains like commercial banking where customer servicing teams manually read, sort, and respond to hundreds of emails each day. These emails often include attachments, unstructured text, and complex request types that make manual triage slow and error-prone. Our solution uses AI to classify request types, extract sub-requests, process attachments, and detect duplicate emails at scale, reducing operational overhead and improving response efficiency.

AgentLLMRAGGPT-4

Mar 2025

Word Dragon

Word Dragon is a vocabulary-building web app where users can search for word meanings and organize them into personalized lists for future learning. It's a simple yet powerful tool to improve and track your vocabulary over time.

Next.jsNode.jsMongoDBExpress

Jun 2024

Resume Match

Resume Match is an AI-powered tool that streamlines the process of matching resumes to job descriptions, solving inefficiencies in traditional recruiting platforms. By leveraging GPT-3.5 and vector search, it provides recruiters and candidates with smart, scalable job-resume pairing and visual insights.

Azure OpenAIGPT-3.5Vector SearchUMAP

Apr 2024

Gym Buddy

Gym Buddy is a fitness web app that helps users find gym exercises tailored to specific body parts, target muscles, and equipment. It's a comprehensive tool for planning and discovering effective workout routines.

Next.jsExpressNode.jsMongoDB

Sep 2023

Gandalf: Harnessing DistilBERT Transformer and BiLSTM for Precise Website Content Classification and Blocking

The Internet has become an essential part of one’s daily life, offering vast amounts of information at one’s fingertips. However, it also presents a significant threat to young web users who may encounter inappropriate or harmful content. To address this issue, a deep learning model, named the “Gandalf Model”, is proposed for accurately identifying the type of content on a website. The Gandalf Model uses a DistilBERT Transformer and BiLSTM to classify websites into ten categories, including NSFW. The model has been trained on a large dataset of over 400,000 data entries, which has been curated and labeled for this purpose. The experimental results demonstrate that the model achieves an accuracy of 80.1%, which is better than many traditional models used in this field. The model protects young web users from inappropriate or harmful content on the Internet thus contributing to the development of effective content filtering and blocking systems. By coupling this model with a proxy server, access to websites belonging to categories as desired by the system administrators is also successfully blocked.

CybersecurityContent BlockerTransformerTensorflow

Jul 2023

PadhAI: AI-based Learning Platform

PadhAI is an innovative Learning Management System (LMS) that leverages cutting-edge technology to redefine the way we learn. Powered by the state-of-the-art GPT-4 and integrated with the dynamic YouTube API, Studcate creates personalized courses autonomously. This groundbreaking project was initiated during the CFG'23 event, hosted by JPMC, marking a significant milestone in the fusion of AI and education.

LangChainGPT-4Next.jsNode.js

Jun 2023

Revealing and Categorizing Weeds in Cotton Plants using XAI and RepVGGPlus

The main focus of this research project is to develop an explainable AI-based deep learning framework for revealing and categorizing of crop weeds in cotton plants. Weed revealing and categorizing is an essential task in precision agriculture as it enables farmers to apply herbicides selectively, reducing the amount of chemicals used and thus minimizing the influence on the atmosphere and human wellbeing. This framework is designed using image pre-processing techniques to crop relevant structures from the images, a deep neural network RepVggPlus to classify the presence of weeds, and finally explainable AI using LIME to elaborate on predicting why these particular results are shown. Compared to existing techniques, our proposed approach offers several advantages such as it can reveal and categorize multiple weed species simultaneously, which is not possible with most existing methods; moreover, our system provides explainable results, allowing farmers to understand how the decision was made and increasing their trust in the system. The outcome is a model that can be used as a tool that can assist farmers and agricultural workers in identifying and removing weeds, leading to increased crop yields and reduced use of herbicides. The framework designed is validated and tested on various different crop fields, to increase its generalization power, and to make sure it will work under different lighting conditions, angles of view and image resolution. This framework uses RepVggPlus which is compared with other deep learning models and proves to be better in terms of accuracy and prediction rate.

Computer VisionExplainable AIMulti-class Classifier

Mar 2023

ChatGPT CLI

ChatGPT CLI is an innovative command-line tool that brings the power of ChatGPT to your fingertips. With features like personas, conversation history, and much more, it's a versatile and dynamic platform for engaging in conversations with AI right from your Terminal.

ChatGPTPython

Mar 2023

Check In/Out

Check In/Out is a versatile application designed to streamline event management, particularly for large gatherings. This application, which made its debut at WomenTechies'23, offers a plethora of useful features to enhance the event experience. From keeping track of total participant counts to monitoring the number of people inside and outside the hall, to facilitating team tracking and detailed event logs, Check In/Out simplifies event logistics and ensures a smoother event execution. This application was used in WomenTechies'23.

Node.jsExpressMongoDBStreamlit

Jan 2023

NotSoCold

NotSoCold is an AI-powered Chrome extension tailored to assist students in discovering off-campus internship opportunities and composing personalized outreach messages. This innovative tool simplifies the internship search process and empowers students to connect with potential employers in a more engaging and effective manner. It doesn't stop there - it also offers a range of additional features, including comprehensive company lists categorized by tags such as "product-based" and "12LPA+", and a Twitter (now X) hiring tweets scraper that displays all relevant hiring-related tweets.

GPT-3ExtensionNode.jsExpress

Jan 2023

Local Communication Without Internet

Transmitting messages using a flashlight instead of internet. A mini project made by the team Orange Crush for the problem statement given by Permissionless.

PythonOpenCV

Dec 2022

OurChatStory.co - #WhatsAppWrapped

OurChatStory.co is a captivating web application designed to provide you with intriguing insights into your WhatsApp chats that you can share with friends and peers. Discover fun facts about your conversations, such as who chats the most, who initiates conversations, when you are most active, and even visualize your chat history with a unique word cloud. Share your "wraps" with friends and explore the fascinating dynamics of your messaging history.

NLTKFastAPINextJSDigitalOcean

Nov 2022

Intrusion Detection System using Neural Networks

This project delves into a comparative analysis of various neural network types employed in Intrusion Detection Systems (IDS). With the goal of enhancing network security and identifying potential threats, we explore the effectiveness of different neural network architectures. In this study, we have conducted a comprehensive comparison among three primary neural network types for intrusion detection: Shallow Neural Network, Deep Neural Network and Convolutional Neural Network.

CybersecurityTensorflowIDS

Nov 2022

Cryptography using Quantum Computing

Attacks against the cryptographic algorithm are one of the many parameters that can be used in a cryptographic assault. The project focuses on how most types of conventional public key encryption are expected to be broken by quantum computing. It highlights how finding quantum-safe algorithms and practices is therefore of utmost importance. Online financial transactions and communications could be targeted by hackers with access to quantum computers if there aren't quantum-safe cryptographic safeguards in place for everything from autonomous vehicles to military hardware. This project, via various modules, focuses on the following topics in particular: current cryptographic schemes (symmetric and asymmetric), quantum computing challenges, differences between quantum and classical computing, quantum algorithms ,affected public key encryption schemes and post quantum cryptography.

CybersecurityQuantum ComputingQiskit

Nov 2022

Automatic Traffic System

The Automatic Traffic System is a cutting-edge solution for the challenges faced by traditional traffic control systems in metro cities of India. Conventional traffic signal timers often lead to inefficiencies due to fixed time intervals for switching traffic directions, regardless of the actual traffic density. Our project leverages the power of the Internet of Things (IoT) and Arduino to create a dynamic traffic control system. By incorporating sound sensors capable of detecting sound wavelengths and frequencies, the system adapts to real-time traffic conditions. For instance, when an ambulance approaches, the traffic signal can be smartly set to green to facilitate its swift journey to the hospital.

IOTArduino

Nov 2022

Blood Bank Monitoring System

The Blood Bank Monitoring System is an IoT device created using NodeMCU and DHT11 sensors, designed to safeguard the integrity of blood storage by continuously monitoring temperature levels. In the wake of the COVID-19 pandemic, the medical field faced numerous challenges, including limited staff availability and the need for innovative solutions to maintain the safety and quality of blood supplies. Our project addresses these challenges by introducing an IoT-based system that keeps a vigilant eye on blood temperatures. Blood, being highly sensitive to temperature fluctuations, can become contaminated if not stored at the right conditions. To prevent this, we employ a blood-like liquid for testing, simulating the real-world blood storage scenario. Temperature data from these simulations are periodically transmitted to a central server, enabling timely notifications and actions to maintain the quality of stored blood.

IOTNodeMCUDHT11

Oct 2022

Multi-Factor Authentication using Keystroke Analysis

This project serves as a Proof of Concept (POC) application designed to implement multi-factor authentication through the unique method of keystroke analysis using n-grams. Multi-factor authentication (MFA) is a critical component of cybersecurity, enhancing the security of digital systems by requiring multiple forms of verification. In this innovative approach, the project focuses on keystroke analysis as one of the factors for authentication. It assesses the timing and rhythm of keystrokes (n-grams) to create a distinctive user profile. By analyzing keystroke patterns, the system can effectively determine the authenticity of the user, adding an extra layer of security.

CybersecurityPythonKeylogger

Sep 2022

Mexpense - Paid Mess Credits Tracker

A paid mess tracker app to efficiently manage and monitor daily mess credits.

Android

Aug 2022

MOM Bot

MOM Bot is a Discord bot that utilizes the T5 Transformer to simplify the process of creating Meeting Minutes (MOM) from the transcript of a meeting. Meeting minutes are essential for documenting discussions, decisions, and action items, but creating them manually can be time-consuming and tedious. This innovative bot automates the generation of MOM by extracting key insights and content from the meeting transcript. By employing state-of-the-art Natural Language Processing (NLP) technologies, it ensures that valuable information is captured accurately and efficiently.

PyTorchTransformersNLPFastAPI

May 2022

Clumos: Club Management Portal

Clumos is your one-stop solution for effective club management. It's designed to empower club boards with the tools they need to efficiently oversee and organize club activities. From tracking projects and events to communicating with club members, Clumos offers a comprehensive suite of features to simplify club operations with features like project and events tracker, an announcement system, video calling capabilities, a milestone tracker, messaging system, and much more. Clumos brings together all the essential functionalities to streamline club administration and improve member engagement.

ReactNode.jsMySQLWebRTC

Apr 2022

Superstore DV Dashboard

The Superstore DV Dashboard project focuses on the development of visual idioms to provide a deep understanding of sales trends within various provinces of the United States of America. In the world of retail and commerce, sales data analysis is pivotal for making informed decisions, optimizing operations, and enhancing profitability. This project introduces a powerful tool that leverages data visualization techniques to decipher complex sales trends and patterns. By utilizing the R programming language and custom dashboards, it offers a comprehensive and intuitive way to explore sales data across different provinces of the USA.

Data VisualisationRDashboard

Dec 2021

Falling Rocks: Game

Falling Rocks: Game is an exhilarating adventure crafted using JavaFX and multithreading. In this action-packed game, players are thrust into a world of peril as they must navigate through a treacherous environment, avoiding collision with an incessant downpour of falling rocks. With a dynamic blend of Java, multithreading, and JavaFX, this project delivers a thrilling gaming experience, where quick reflexes and strategic planning are essential for survival.

JavaMultithreadingJavaFX

Oct 2021

Auto Grader

Auto Grader is a cutting-edge solution that streamlines the grading process by leveraging a deep learning pipeline. This innovative project automates the evaluation of answer sheets, making use of Computer Vision (CV) and Optical Character Recognition (OCR) to read and extract answers. It then employs BERT embeddings for similarity analysis, allowing for efficient and objective grading. It significantly reduces the time and effort required for manual grading while ensuring accuracy and consistency in the evaluation process. By combining CV, NLP, Tensorflow, FastAPI, and Docker, this project offers a versatile and robust solution for educators and institutions.

CVNLPTensorflowFastAPI

Aug 2021

VITTY: VIT Timetable App

VITTY is an Android app designed to make the lives of VIT (Vellore Institute of Technology) students more convenient. With the app, students can effortlessly access their timetables, receive timely notifications, and efficiently navigate to their classes. This project offers a simplified interface, home screen widgets, and the added convenience of keeping track of their class schedules. Catering the needs of over 25,000 users, VITTY has made class management more accessible and user-friendly, ensuring that students are always well-informed about their upcoming classes.

AndroidFirebase

Jul 2021

Cadence

Cadence is an Android app that leverages machine learning to choose the perfect wake-up song based on your alarm type and mood. It integrates seamlessly with Spotify to enhance your mornings with personalized music recommendations.

AndroidSpotify SDKMachine Learning

Jun 2021

WhatsAppThem

A simple PWA that you can use to send WhatsApp messages without saving the number.

JavaScriptHTML/CSS

May 2021

TRIEPAD

A notepad with features like context-aware auto-complete and dictionary made using the trie data structure.

Data StructureAlgorithmTrie

May 2021

Odin

A Discord bot built with Python and TerraForm to create cloud-based virtual machines on AWS, Digital Ocean, and others.

DevOpsCloudTerraformAutomation

Apr 2021

HypeLinks API

An API which can classify website URLs into 10 different categories. The API was made using FastAPI (deployed on Azure) and the model was made in Tensorflow using DistilBert Transformer from Hugging Face.

TensorflowTransformersNLPFastAPI

Mar 2021

WomenTechies'21

An Android app for the event WomenTechies'21. The app can send realtime notifications, show in-app announcements, Instagram highlights and timeline of the event, and other info.

AndroidFirebase

Dec 2020

Sentival

A chrome extension that helps the readers to understand the underlined subconscious biases given by the author in the article and gives an unbiased analysis.

ExtensionNLPNode.js