Gnanasankari's Portfolio

my work

Projects I Did

Gnanasankari's Portfolio Website

Developed a personal portfolio website as my first project, using HTML, CSS, and JavaScript to showcase skills and projects. Gained hands on experience in responsive design, UI components, and project deployment

To-Do Task List

A simple and intuitive task management app to add, mark, and organize daily tasks. Designed to practice DOM manipulation, event handling, and responsive UI

Scientific Calculator Web Application

Designed and developed a scientific calculator using HTML, CSS, and JavaScript, focusing on accurate mathematical operations, clean UI design, and intuitive user interaction. Deployed end-to-end using GitHub Pages.

AI Task Automation Agent

AI Task Automation Platform is an intelligent productivity system that unifies Google Calendar, to-do lists, and reminders into a single dashboard for seamless daily planning. This project is an extended version of my initial to-do list application, enhanced to handle real-world scheduling needs more effectively. It features chatbot integration that allows users to manage tasks and schedules using natural language interactions. The system focuses on reducing manual effort and helping users stay organized through smart automation.

Tamil E-Newspaper Website

Designed and developed a sample Tamil e-newspaper website to present regional news content in a clean and readable digital format. Built using HTML, CSS, and JavaScript, and deployed end-to-end using GitHub Pages. The project focuses on content structure, UI readability, and basic frontend best practices. Designed and developed a sample Tamil e-newspaper website to present regional news content in a clean and readable digital format. Built using HTML, CSS, and JavaScript, and deployed end-to-end using GitHub Pages. The project focuses on content structure, UI readability, and basic frontend best practices.

Phoenix Dial : Smart Sales Call Tracking System

Phoenix Dial is a sales call tracking system designed to provide end-to-end visibility into call activities and performance metrics. It supports lead assignment, call logging, outcome tracking, and structured follow-up management. Managers gain actionable insights into call duration, response quality, and individual performance trends. The platform was built as a practical, real world solution to improve sales efficiency and operational transparency. It enforces accountability through centralized data capture and analytics-driven reporting. The system was implemented using the MERN stack, with a scalable backend, RESTful APIs, and a responsive frontend architecture.

Stock Sheet Management System

A structured system to track, organize, and manage stock data efficiently, emphasizing clean data handling and real world operational workflows. Built to enhance accuracy, maintainability, and clarity in day-to-day stock management processes.

UIDAI Hackathon 2026 – Aadhaar Data Analytics

Created a structured analytics system to explore Aadhaar enrollment patterns, focusing on data accuracy, clarity and real world applicability.

Journal Automation : Personal Accounting Workflow System

Built a Journal Automation system to manage multi account transactions with real-time Money In/Out tracking and dynamic opening/closing balance calculation. Designed with a backend first approach focusing on clean data modeling, transaction consistency, and scalable accounting workflows.

Appointment Scheduler using REST APIs and MySQL

Developed an appointment booking and scheduling system that manages users, time slots and appointments through REST APIs. The system uses a relational database to store and organize scheduling data while maintaining proper relationships between users, available slots, and bookings. It supports creating, updating, and managing appointments while ensuring consistent and structured data handling. The project demonstrates how well designed APIs and structured database models help build reliable backend systems for applications that require appointment and schedule management.

Smart Healthcare Assistant for Ear and Eye Diagnosis

Developed an AI-powered healthcare web platform during VMedithon that detects potential ear and eye conditions from uploaded images, helping doctors with faster and more accurate preliminary analysis. The platform acts as a smart assistant by generating concise summaries of prescriptions, reducing manual effort and improving clarity for patients. It also includes features such as online appointment booking for seamless doctor consultations and an integrated chatbot to handle basic patient queries and provide guidance, making the overall healthcare experience more efficient and accessible.

Kids Live Tracker System

Developed a real time child tracking system to enhance safety by continuously monitoring location and movement using GPS and IoT technologies. Enabled accurate, low latency location updates with seamless communication between device and application. Implemented geofencing and anomaly detection to trigger instant alerts for boundary breaches or unusual activities. Designed a user friendly interface for easy monitoring, with a scalable backend to support multiple users and devices.

VoiceIT Student Club Website (Frontend Contribution)

Contributed to the VoiceIT Student Club website as part of the frontend team. Worked on building and improving UI components using HTML, CSS, and JavaScript. Collaborated in a team environment to enhance usability and ensure a responsive design.

Smart Energy Harvesting System (Simulink Based)

Designed and simulated an energy harvesting system using MATLAB Simulink to efficiently convert and store power. Integrated components like a rectifier, buck converter, and battery to ensure stable DC output and continuous charging.

UAV Swarm Security Framework (AI + Blockchain Based)

Designed and developed a real-time UAV swarm security system integrating AI-driven intrusion detection with blockchain-based secure communication. The system uses machine learning models to detect anomalies in network traffic and ensures tamper-proof data exchange using Hyperledger Fabric with AES-256 and ECC encryption.

WebRTC Video Call & Live Chat Web Application

Built a browser based real time communication platform with WebRTC video calling, live chat, and file sharing. Supports room-based peer-to-peer sessions with low latency using Node.js, Socket.IO, and JavaScript.

Voice2Action (V2A) – Civic Task Routing Platform

A civic tech platform that turns community issues into micro employment opportunities. Citizens report problems, tasks are routed to authorities and local workers, with transparent tracking and verified digital payments.

Cloud Native VTOP Platform with DevOps, MLOps & AI

I built a cloud native VTOP platform integrating DevOps, MLOps, AI, and modern software engineering practices. It utilizes Java Spring Boot, React, Docker, Kubernetes, Helm, CI/CD pipelines, and MLflow for scalable deployment and AI powered student analytics.

Patch Paladin 2.0

I designed and developed Patch Paladin 2.0, an intelligent cybersecurity solution for automated vulnerability scanning and patch management. It integrates CVE databases to identify risks in Windows systems, detect outdated applications, and deploy real time updates. The platform features an interactive dashboard for scan history and real time security analysis to simplify vulnerability assessment and strengthen endpoint protection.

SDN Traffic Prioritization System

I designed and developed an SDN based Traffic Prioritization System that intelligently classifies and manages network traffic to improve bandwidth utilization and Quality of Service. It features a real time classification engine for varying priority flows across a simulated topology with centralized controllers and OpenFlow switches. The system includes an interactive dashboard to visualize network topology, traffic statistics, and packet drops while integrating REST APIs to dynamically configure traffic priorities.

TASA CodeCraft Task Management Board

I designed and developed a modern Kanban based Task Management Board for the TASA CodeCraft platform to simplify project planning and progress tracking. It features multi user workspaces, real time task status tracking, and dynamic counters within a clean, responsive interface. The application serves as an internal solution enabling efficient collaboration, transparent project tracking, and streamlined workflow management for teams.

Machine Unlearning in Keyword Spotting Systems

I worked on a research project investigating techniques to remove the influence of specific training samples from deep learning models without full retraining. I evaluated different unlearning approaches based on forgetting effectiveness, model utility, and computational efficiency, directly contributing to privacy preserving AI research.

Billing and Financial Automation Platform

I developed an end to end billing and financial automation platform that integrates POS, Zoho Books, Google Sheets and banking workflows. The platform automates invoice synchronization, payslip generation, bank statement processing, GST reporting, daily expense tracking and sales analysis. It includes a centralized dashboard to monitor automation status, pending tasks and generated reports, reducing manual work and improving visibility into daily financial operations.

experience

Professional Journey

Virtual Internship with Samsung Prism

Feb 2026 – Jun 2026

Samsung R&D Institute India · Bangalore · Remote

Worked on the Samsung PRISM Virtual Internship project titled "Machine Unlearning in Keyword Spotting Systems," focusing on developing privacy preserving AI solutions for speech recognition.
Researched, implemented, and evaluated multiple machine unlearning techniques to selectively remove the influence of specific training data while preserving the model's performance on retained data.
Conducted extensive experimentation and comparative analysis of different unlearning approaches, measuring their effectiveness based on forgetting accuracy, retained model utility and computational efficiency.
Collaborated closely with mentors throughout the project, gaining hands on experience in applied AI research, machine learning experimentation and developing privacy aware intelligent systems.

Career Break

Jun 2020 – May 2025

Full-time Parenting · Gudiyattam, Tamil Nadu

Focused on family responsibilities and childcare.
Continued self learning in programming and software development.
Prepared for higher studies and career re-entry.

Software Engineer

Jun 2019 – May 2020

Odessa Technologies · Full-time · Bengaluru

Worked in Security Bank Project.
Developed most of the client requirements and contributed nearly 40% of the development work in that project.
Some of the works include Non Performing Assets module, Multiple Job Creations as per the Client Request, Handling Client Issues etc.
In annual performance appraisal, I got a SUPER STAR RATING which opened up another opportunity to work as a developer in TOYOTA Project as well simultaneously.
Mentored junior developers, conducted code reviews for them and helped them to improve their code quality and ensure adherence to best practices.

Software Intern

Jun 2018 – Jul 2018

Zoho · Chennai, Tamil Nadu

Developed web application to store student details in a database from various colleges.
Integrated that web application with the DBMS.
Learned JAVA Programming Language and how to integrate JAVA in real world applications.