Ayon Sen

MACHINE LEARNING ENGINEER & FULL-STACK DEVELOPER

TURNINGIDEASINTOIMPACT.

OBSESSED WITH SOLVING PROBLEMS THROUGH AI.

I'm a Computer Science graduate passionate about building real-world AI systems— from Natural Language Processing to Computer Vision and Explainable AI. I bring ideas to life with Python, PyTorch, and full-stack tools like React and Node. Let's make technology meaningful.

Scroll to explore

EXPERTISE

TECHNICALPROWESS

Combining cutting-edge technologies with creative problem-solving to build scalable, user-centric solutions that make an impact.

TensorFlow

AI & Machine Learning

PyTorch

AI & Machine Learning

Scikit-learn

AI & Machine Learning

Keras NLP

AI & Machine Learning

Neural Networks

AI & Machine Learning

Computer Vision

AI & Machine Learning

LLM Integration

AI & Machine Learning

Prompt Engineering

AI & Machine Learning

AI Agents

AI & Machine Learning

OpenAI APIs

AI & Machine Learning

Roboflow

AI & Machine Learning

React.js

Frontend Development

Next.js

Frontend Development

JavaScript

Frontend Development

TypeScript

Frontend Development

HTML5/CSS3

Frontend Development

Tailwind CSS

Frontend Development

Redux

Frontend Development

SEO

Frontend Development

React Native

Mobile Development

Expo

Mobile Development

Android Development

Mobile Development

Python

Backend Development

Django

Backend Development

Flask

Backend Development

Node.js

Backend Development

Express

Backend Development

PHP-Laravel

Backend Development

REST APIs

Backend Development

Prisma

Backend Development

PostgreSQL

Cloud & DevOps

MongoDB

Cloud & DevOps

MySQL

Cloud & DevOps

Firebase

Cloud & DevOps

Azure

Cloud & DevOps

Railway

Cloud & DevOps

Render

Cloud & DevOps

Vercel

Cloud & DevOps

Git

Cloud & DevOps

GitHub

Cloud & DevOps

GitLab

Cloud & DevOps

Figma

Design & UI/UX

Canva

Design & UI/UX

Pandas

AI & Machine Learning

NumPy

AI & Machine Learning

R (tidyverse)

AI & Machine Learning

MATLAB

AI & Machine Learning

Power BI

AI & Machine Learning

Tableau

AI & Machine Learning

Microsoft Excel

AI & Machine Learning

FEATURED WORK

SelectedProjects

Tong - The Social Media Web App

Tong is a community-driven social media web app inspired by the vibrant culture of Bangladeshi tea stalls (tong). It enables users to connect, debate, and share stories in a lively, interactive environment.

Tong - The Social Media Web App

TypeScriptReactNode.jsWebSocketsMongoDB
Bone Tumor Prediction

Published research in Springer on bone tumor prediction using machine learning techniques. The study focuses on developing accurate prediction models for bone tumor classification and early detection.

Bone Tumor Prediction

Machine LearningMedical ImagingPythonResearchData AnalysisDeep Learning
Bangla News Event Detection

Research work on event detection in Bangla news text using advanced NLP techniques. Implemented a full pipeline with text preprocessing, transformer-based embeddings, and deep learning models.

Bangla News Event Detection

NLPTransformersBiLSTMFastTextPythonResearch
Bangla Braille Recognition

Research project on recognizing Bangla Braille characters using computer vision and deep learning. Developed a CNN-based system achieving 94% mAP@50 for real-time character recognition.

Bangla Braille Recognition

Computer VisionYOLOv8CNNPythonResearchAccessibility
AI Agent Chatbot

An advanced AI chatbot built using LangGraph, FastAPI, and Streamlit. Features intelligent agent-based interactions and natural language processing capabilities.

AI Agent Chatbot

PythonLangGraphFastAPIStreamlitAINLP
Football Field Analytics

An intelligent system for analyzing football matches through video using Machine Learning and OpenCV. Combines deep learning-based object detection (YOLOv5) with traditional computer vision techniques.

Football Field Analytics

PythonOpenCVYOLOv5Machine LearningComputer VisionJupyter Notebooks
Ghotona Chitro

A web application that detects events from Bangla text sources using advanced NLP techniques and machine learning models.

Ghotona Chitro

JavaScriptNLPMachine LearningWeb Development
Opal

A video messaging tool that lets you quickly record your screen, voice, and face to share information faster than typing. Perfect for async communication, tutorials, product demos, and team updates.

Opal

TypeScriptReactVideo ProcessingWebRTC
Machine Learning Assignments

A collection of Machine Learning practice assignments and implementations covering various ML concepts and techniques.

Machine Learning Assignments

PythonMachine LearningJupyter NotebooksData Analysis
Samayrizer

An efficient tool for summarizing PDF documents quickly and accurately using advanced NLP techniques.

Samayrizer

PythonNLPPDF ProcessingText Summarization
Bangla Article Dataset

A curated collection of Bangla news articles for NLP research and development, supporting various natural language processing tasks.

Bangla Article Dataset

PythonNLPData CollectionBangla Language Processing
RestaurantSys

A comprehensive restaurant management system with features for menu management, order processing, and customer service.

RestaurantSys

PythonWeb DevelopmentDatabase ManagementSystem Design

PROFESSIONAL JOURNEY

Experience&WorkHistory

A timeline of my professional growth, key projects, and achievements in software development and innovation.

2024 - Present

Undergraduate Researcher – NLP

East West University

Designed and implemented a full NLP pipeline for event detection in Bangla news text using both traditional and deep learning models.

  • Built preprocessing pipeline with text cleaning, tokenization, TF-IDF, FastText, and transformer-based embeddings
  • Developed classifiers including SVM, Random Forest, and BiLSTM with attention achieving over 90% accuracy
  • Enhanced performance using Named-Entity Recognition and dependency parsing
  • Developed a full-stack web app (React.js frontend + Flask backend) for real-time classification and visualization
  • Conducted robustness evaluation across multiple Bangla news sources
PythonFlaskReact.jsBiLSTMFastTextTF-IDFTransformersScikit-learnNERPandas
2024 - Present

Undergraduate Researcher – Computer Vision

East West University

Developed a CNN-based system for recognizing Bangla Braille characters from dot-pattern images to support accessibility tools for the visually impaired.

  • Manually compiled and labeled a dataset of 30,000 images for 50 Braille characters
  • Applied data augmentation and labeling via Roboflow
  • Evaluated multiple models including YOLOv8n, YOLOv8s, and Faster R-CNN
  • Achieved 94% mAP@50 and 80% mAP@50–95 with YOLOv8s, enabling real-time usage
  • Highlighted trade-offs between model speed and detection granularity
PythonYOLOv8Faster R-CNNRoboflowOpenCVCNNPyTorchImage Augmentation

GET IN TOUCH

Let'sCreateSomethingAmazing

Have a project in mind or want to explore possibilities? I'm always open to discussing new opportunities and innovative ideas.