Embedded Files
This page contains brief overviews of all the projects that I have done during my college tenure.
Table of Contents
Table of Contents
Online Learning Management Platform with GenAI Tools
Online Learning Management Platform with GenAI Tools
StudyBuddy, where Artificial Intelligence enhances the learning experience. Engage with course materials, submit various assignments, and get AI-powered support.
Features
Features
Course Content Access: Read and interact with diverse educational materials.
Assignment Submissions: Submit MCQs, MSQs, numeric problems, and coding tasks.
AI Assistance: Get material summaries, contextual hints, and interactive document chat.
Coding Support: Receive hints and guidance for coding assignments. it various assignments, and get AI-powered support.
Technologies Used
Technologies Used
Python 3: The primary programming language for the application.
FastAPI: Used for building the backend of the application, providing fast and efficient API endpoints with built-in validation and documentation.
React: A JavaScript library for building user interfaces, used in the frontend of the application.
MongoDB: A NoSQL database used to store and manage application data.
LangChain: A framework for building applications with LLMs (Large Language Models)
Smart Attendance System
Smart Attendance System
(ongoing)
The Smart Attendance System is an advanced solution that seamlessly integrates geolocation, face recognition, and Bluetooth proximity to simplify attendance management. Features like real-time tracking, automated timers, and detailed reportsmake recording attendance easier than ever.
Features
Features
Location-Based Tracking: Automated detection of entry and exit using geofencing to ensure accurate attendance logs.
Bluetooth Proximity Monitoring: Tracks employee proximity to their assigned workspace, triggering timers and notifications for efficiency.
Facial Recognition: Advanced authentication via face recognition ensures security and reliability.
End-of-Day Summaries: Attendance data consolidated and sent to the backend for detailed analysis.
Attendance Reports: Comprehensive reports with visually appealing graphs and summaries for individuals and teams.
Manager Tools: Manage team attendance with features like manual time adjustments, attendance correction approvals, and action logs.
Technologies Used
Technologies Used
Frontend: React Native for an intuitive and dynamic mobile app interface.
Backend: Flask for robust API development and MongoDB for efficient database management. Google Cloud Platform for hosting
Face Recognition: Azure Face API for advanced and reliable facial recognition.
Geofencing: Haversine distance calculation to detect location-based entry and exit.
Bluetooth Proximity: RSSI-based tracking for real-time proximity detection.
Visualizations: React charting libraries for interactive and insightful data representation.
AI-Powered Disaster Relief Assistant
AI-Powered Disaster Relief Assistant
(ongoing)
This is a robust platform designed to streamline disaster response efforts. It leverages AI for real-time geocoding, social media analysis, and local language translation to provide critical information and support during emergencies.
Features
Features
Address Geocoding: Converts location data into precise geographical coordinates to improve response accuracy.
Local Language Translation: Provides multilingual support for seamless communication in disaster-affected regions.
Chatbot Integration: Offers 24/7 assistance for queries and guidance during emergencies.
User-Friendly Dashboard: Displays analyzed data and actionable insights in an intuitive interface.
Technologies Used
Technologies Used
Backend: Flask
Geocoding: Azure Maps
AI Models: Azure OpenAI Service
Language Services: Azure Language Service
Online Grocery Store (v2)
Online Grocery Store (v2)
This project includes most features of the "Online Grocery Store (v1)", plus some extra features. Also, the frontend and backend are separately done, unlike the previous project. Frontend using Vue.js and Backend using Flask.
Extra features:
Extra features:
User-end
User-end
Authentication: With JSON Web Tokens.
Reminders and Reports: Daily reminders and monthly reports will be mailed to users
RBAC has been implemented. So 2 other roles are Store Manager and Admin.
Store Manager and Admin-end
Store Manager and Admin-end
Product Management: Admins and Managers are granted the authority to not only add new products to the system but also update existing product details, such as prices, quantities, manufacturing dates, and expiry dates.
Download Report: Admins and Managers can download sales reports in CSV format.
Some more features
Some more features
Admins can promote users to store managers.
Caching on redis database
Pre & Post-COVID Analysis of RCSB East Associates
Pre & Post-COVID Analysis of RCSB East Associates
Made a comprehensive "Pre and Post-COVID Analysis" for RCSB East Associates, a Godrej dealership, utilizing advanced data analytics, visualizations, and strategic insights to optimize sales and profit, fostering adaptability and resilience in a dynamic market landscape.
I want to thank the CEOs of RCSB Associates, Mr. Saugata Bhattacharyya and Mr. Rana Chandra, for providing access to the company data, allowing me to perform analysis for this project, and permitting me to share my findings.
Taxi Fare Prediction
Taxi Fare Prediction
Machine Learning Project
Machine Learning Project
This project focuses on solving a regression problem using advanced machine learning techniques. I implemented multiple models, including Bagging Regressor and XGBoost, and evaluated their performance using R-squared and Mean Squared Error (MSE) on both training and test datasets. To ensure model robustness, I utilized cross-validation, comparing the mean and standard deviation of the scores across different folds. Visualizations were generated to highlight performance variations among models, making the results easier to interpret. The final model's predictions were exported as a CSV file, completing the full machine learning pipeline from data preprocessing to model deployment.
Online Grocery Store (v1)
Online Grocery Store (v1)
User-end
User-end
Authorization and Security: The users and manager can register and log in securely. The passwords are hashed using the SHA-256 cryptographic hashing algorithm before being stored in the database.
Searching and Filtering: Users can search for products using keywords and apply filters based on price, manufacture date, and expiry date. The project facilitates a user-friendly search experience to help users find products efficiently.
Shopping Cart: Users can add products to their shopping carts, update quantities, and remove items. The shopping cart functionality is designed to provide real-time feedback to users and ensure a smooth checkout process.
Coupon Application: Users can apply coupon codes during the checkout process to receive discounts on their orders. The project validates coupon codes and adjusts the total price accordingly.
Order Processing and Confirmation: Once users are satisfied with their shopping cart, they can proceed to checkout. The checkout process involves capturing user information, validating the order, and recording the order in the database.
Manager-end
Manager-end
Product Management: Managers are granted the authority to not only add new products to the system but also update existing product details, such as prices, quantities, manufacturing dates, and expiry dates.
Category and Coupon Code Management: Although product categories can be automatically added and deleted during product management, Managers may create and delete categories. Also, Managers can add, remove, and monitor the use of coupons.
Chess-Piece Identification
Chess-Piece Identification
Deep Learning using Keras
Deep Learning using Keras
This project is about classifying images of chess pieces. The dataset contains images of six types of chess pieces: Bishop, King, Knight, Pawn, Queen, and Rook. After performing some exploratory data analysis, a sequential model is built using convolutional layers for feature extraction and dense layers for classification. The model is then trained and evaluated, achieving a certain level of accuracy on a held-out validation set.
Hackspire Official Website
Hackspire Official Website
Hackspire is an initiative by the FIEM-ACM Student Chapter.
As part of the development team, I contributed to building this site using NextJS, ensuring a seamless and dynamic user experience for our hackathon participants. The website provides all the necessary information about the event, registration details, schedule, and resources for participants.
Task Management Telegram Bot
Task Management Telegram Bot
Worked on this project during my internship at Simtrak Solutions
This bot streamlines task management within teams by leveraging Telegram's simplicity and accessibility. Designed for both regular users and managers, it features an intuitive and visually appealing interface that makes task tracking and delegation seamless.
Features
Features
Task Groups: Assign and manage task groups containing multiple tasks for individual or shared users.
Real-Time Updates: Synchronizes task status instantly across all users sharing the same task group.
Daily Reminders: Notifies users of their tasks every day, making it easy to stay on track.
Manager Controls: Enables managers to assign/retract task groups and monitor task progress effectively.
Activity Logs: Maintains a transparent history of all task-related activities for users and managers.
Role Management: Allows the promotion of users to managers with delegation privileges
Technologies Used:
Technologies Used:
Telegram Bot API: Used for creating and managing the bot
Python: Primary programming language for bot logic and communication with Telegram.
PostgreSQL: Database solution for storing user data, task groups, and activity logs.
Opstream
Opstream
Worked on this project during my internship at Simtrak Solutions
Opstream is a cutting-edge platform designed to revolutionize supply chain communication by providing real-time, actionable insights to stakeholders. It addresses the challenges of delayed information and lack of transparency in supply chains, ensuring that all parties are informed and can make timely decisions.
Features
Features
User-Friendly Interface: Designed with a focus on usability, ensuring that users can navigate the platform effortlessly and access the information they need.
Auto-chaining of messages: Linking messages automatically, based on the context of the messages, thus maintaining.
Message Highlighting: Highlighting messages that belong to a particular chain.
Technologies Used:
Technologies Used:
Next.js: Used for developing the frontend, enabling server-side rendering and a seamless user experience.
FastAPI: Employed for building high-performance APIs, ensuring efficient data handling and proper documentation.
OpenAI API: Predicting the context and status of messages.
PostgreSQL: Utilized for reliable and scalable data storage.
Page updated
Report abuse