Hey there, I'm Mugil

• AI/ML/DL Researcher • Full Stack Developer

Final-year Computer Science undergrad at SSNCE
Dedicated to building scalable systems that solve real-world problems with precision.

Get in Touch View Projects View ResumeOpens in Google Drive

ABOUT ME

I'm Mugil — a final-year CSE undergraduate at SSN College of Engineering. I'm passionate about building and experimenting with Machine Learning and Deep Learning models to tackle real-world problems. I also enjoy developing full-stack web applications using the latest modern frameworks.

I'm driven by curiosity, growth, and collaboration — always eager to learn, evolve, and connect with people who share the same energy.

Experience & Activities

IIT Roorkee

Research Intern

Jun 2025 – Present

Exploring deep learning models for Alzheimer’s diagnosis using PyTorch, U-Net, VoxCNN, and ResNet18. Developed a unique PyTorch-based approach within 6 weeks, integrating GANs and LRP-Z for counterfactual explainability using FSL tools

IEEE EMBS SSN

Web Development Head

Jun 2025 – Present

Currently leading development of the society’s website planned to deploy using TypeScript, Tailwind CSS, and Firebase. Aiming to streamline event hosting, smooth UI experience and improve participant engagement

AI Student Chapter SSN

PR/Sponsorship Core

Mar 2025 – Present

Involved in outreach across Gmail campaigns and Instagram to maximize event reach and visibility. Supported promotional efforts and maintained public engagement

$Event Head - Exlog$

SSN Math Club

Event Head - Exlog

Jan 2025 – Feb 2025

Planned and led the "Rapid Fire" event with more than 70 participants. Handled everything from quiz preparation to anchoring the event, receiving strong feedback for smooth coordination

ACE, SSN

Marketing & Logistics Core

Aug 2024 – Sep 2024

Assisted in budget planning, logistics, and student-focused marketing strategies for Invente ’24. Helped improve engagement and ensured smooth material procurement

SSN MUN Society

General Administrator - Logistics

Jun 2024 – Aug 2024

Managed merchandise and delegate kits for the annual MUN. Contributed to the team by effective planning and vendor coordination

Publications

Detecting AI-Written Reviews for Consumer Trust

ACL · May 1, 2025

Published in DravidianLangTech@NAACL 2025, this paper addresses the rising challenge of AI-generated fake reviews in regional languages. Built a multilingual classification framework using mBERT to distinguish synthetic vs human-authored reviews in Tamil and Malayalam. Achieved F1-Score of 0.90 (10^th) in Tamil and 0.68 (28^th) in Malayalam.

Multi-Label Emotion Classification Using Fine-Tuned Roberta-Large Transformer

SemEval-2025 · Accepted

Part of SemEval-2025 Task 11, this paper explores fine-tuned RoBERTa-large models for nuanced emotion detection and regression. Our system ranked 31^st in Macro F1 (0.75) and 17^th in Pearson (0.75).

TECH STACK HELIX

Tech Categories

Languages

Frameworks

Database

DevOps

Design

Technology Stack

Languages

Python

Java

JavaScript

TypeScript

Dart

Assembly

HTML5

CSS3

Frameworks

Flutter

React

Next.js

NodeJS

Express.js

Redux

TailwindCSS

ML

TensorFlow

PyTorch

scikit-learn

OpenCV

Keras

Pandas

NumPy

OmniXAI

Power BI

Database

MySQL

SQLite

MongoDB

Firebase

DevOps

Google Cloud

Docker

Git

GitHub

Design

Figma

"Building one app at a time"

FEATURED PROJECTS

Bankbee

In Progress

A finance management dashboard with modern UI/UX, built using TypeScript, React, and TailwindCSS for professional-grade personal finance insights.

TypeScriptNext.jsTailwindCSSFigma

Credit Card Fraud Detection API

End-to-end ML pipeline for fraud detection with PCA, SMOTE, Flask API, Docker containers, and real-time monitoring using Prometheus + Grafana.

FlaskDockerPrometheusMLGrafana

InternTrack

Full-stack internship tracking app for institutions, including role-based login, OCR for document parsing, and Google Sheets + Drive integration.

ReactFirebaseOCRMongoDBNode.js

IMDB Sentiment Analysis

Sentiment classifier using TF-IDF, 10+ ML models, ensemble voting, and hyperparameter tuning with detailed evaluation on 50K IMDB reviews, deployed in GCP.

PythonMLGCPEnsemble Models

LEAR Framework on MNIST

Implements counterfactual explanations using LEAR framework with encoder-decoder-discriminator architecture on handwritten digit classification.

PyTorchXAI3D CNNsGANs

Defend the Nexus Game

Tower defense game using A* for user-AI and Simulated Annealing for AI-AI, with grid rendering, turret tracking, and wave adaptation.

PythonPygameRaylibGame AI

View More on GitHub→

GET IN TOUCH

Open to internships, collaborations, projects or just talking shop about ML and Fintech topics

mugilkrishna.d.u@gmail.com GitHub LinkedIn Instagram

Chennai