THIS PORTFOLIO IS A RECORD OF THE PROJECTS I HAVE WORKED ON WHILE LEARNING MACHINE LEARNING, AI SYSTEMS, AND SOFTWARE DEVELOPMENT.

View Resume

.About

Highlights

B.Tech Computer Science and Engineering student specializing in Computer Networking at SRMIST, building toward full-stack AI/ML engineering.
AI/ML Research Intern on a Network Digital Twin Project, exploring 6G-oriented digital twins, SLA-breach prediction, explainability, and proactive network-control ideas.
Presented the research work at Nokia Campus Connect 2025, earning third place, and earned first place in a section-level AutoCAD competition.
Built the Source-Cited RAG Assistant to make approved portfolio material easier to explore with validated citations and approved diagrams.
Growing practical MLOps skills alongside Python, Flask, React, data visualization, model evaluation, and explainable ML work.

Skills

Python
Machine Learning
Deep Learning
Flask
HTML
CSS
JavaScript
Data Visualization
MySQL/PostgreSQL
Git
GitHub
Notion
Jupyter
VS Code
Ditto
Docker
Kubernetes

Download Resume

.Projects

Use the Unroll button to explore each project in more detail and access supporting links.

Click the deployed project cards to see live demos of these models in action.

About My Work

My work is driven by a simple aim: turning complex ideas into systems people can actually use and understand. I enjoy starting close to the fundamentals, then carrying that care into research, prototypes, and practical tools. Across machine learning, network digital twins, and source-grounded AI, I focus on asking clear questions, testing assumptions, and communicating the result honestly. Some projects are academic explorations; others are working demos. Together, they reflect how I learn best: patiently, hands-on, and with attention to the details that make a technical project feel useful rather than merely finished.

←

→

Research paper for detailed info

5G Handover with
Stability-Aware ML

An academic ML and simulation project that explores whether recent network KPIs can support earlier, more stable handover recommendations. It combines offline training, a Flask inference API, and an interactive browser simulation rather than controlling a live network.

Decision-tree classification study visualization

Cardio Risk Predictor

An exploratory cardiovascular classification project built from clinical and lifestyle records. I implemented a decision tree from scratch to learn preprocessing, evaluation, and data visualization; it is not medical advice or a clinical tool.

Fleet smart vehicle digital twin visualization

Completion proof

Digital Twin
Fleet Smart Vehicle

A learning prototype for managing vehicle records through a React and TypeScript interface, Flask API, and MongoDB. A separate Python desktop controller sends serial commands to an Arduino horn circuit; the implemented path does not use Ditto or Kubernetes.

Parkinson's Disease Detection

A research pipeline that explores feature-mask search with an attention-based dense network on Parkinson's-related speech-feature records. It uses optional SMOTE resampling, fold-specific checkpoints, and classification reports to compare modeling choices.

#3rd prize winner in NOKIA

Closed Loop Automation

A Python/Jupyter experiment that classifies SLA-violation risk from stored network KPI snapshots. It uses SHAP to inspect a prediction and simulates bounded KPI adjustments in memory as a what-if exercise presented at Nokia Campus Connect 2025.

Fraud transaction classification visualization

Fraud Transaction Detector

An offline machine-learning experiment on a simulated transaction dataset. It engineers transaction features, uses class-weighted gradient boosting, and reports fraud-focused metrics on a held-out test split; it is not a live banking or fraud-prevention system.

AutoML System

An in-progress ML workflow exploring data preparation, model selection, hyperparameter tuning, and model-specific preprocessing. Explainability is planned through SHAP-style analysis; it is not yet presented as a completed production tool.

.Contact

Start a conversation

Open to learning and new opportunities.

.Summary

5G Handover with Stability-Aware ML

Closed Loop Automation

Fleet Smart Vehicle Digital Twin

THIS PORTFOLIO IS A RECORD OF THE PROJECTS I HAVE WORKED ON WHILE LEARNING MACHINE LEARNING, AI SYSTEMS, AND SOFTWARE DEVELOPMENT.

.About

Highlights

.Projects

About My Work

5G Handover with
Stability-Aware ML

Cardio Risk Predictor

Digital Twin
Fleet Smart Vehicle

Parkinson's Disease Detection

Closed Loop Automation

Fraud Transaction Detector

AutoML System

.Contact

Message Me

.Summary

5G Handover with Stability-Aware ML

Closed Loop Automation

Fleet Smart Vehicle Digital Twin

THIS PORTFOLIO IS A RECORD OF THE PROJECTS I HAVE WORKED ON WHILE LEARNING MACHINE LEARNING, AI SYSTEMS, AND SOFTWARE DEVELOPMENT.

.About

Highlights

.Projects

About My Work

5G Handover withStability-Aware ML

Cardio Risk Predictor

Digital TwinFleet Smart Vehicle

Parkinson's Disease Detection

Closed Loop Automation

Fraud Transaction Detector

AutoML System

.Contact

5G Handover with
Stability-Aware ML

Digital Twin
Fleet Smart Vehicle