I'm a full-stack software engineer with a passion for building scalable applications and conducting impactful research. Currently a Jr. Software Engineer at Technext Ltd, where I develop responsive frontends with React and secure backend services with Django.
My expertise spans across modern web technologies, DevOps practices, and computer vision. I'm particularly interested in applying technology to solve real-world problems—from microfinance digitalization to biodiversity monitoring.
With 300+ competitive programming problems solved and multiple published research papers at top-tier conferences, I combine theoretical knowledge with practical engineering skills.
Experience
Jr. Software Engineer (Full-Stack)
Technext LtdSylhet
- –Developed responsive, scalable user interfaces using React (Hooks, TanStack Query), TypeScript, and Tailwind CSS.
- –Implemented secure and efficient backend services with Django (DRF), integrating RESTful APIs and asynchronous tasks with Celery.
- –Streamlined deployment workflows using Docker Compose, Kubernetes, and automated CI/CD pipelines with GitHub Actions.
Jr. SQA Engineer
Kite Games StudioDhaka
- –Performed manual and API testing on multiple mobile and web-based projects.
- –Ensured application reliability and data integrity by validating REST endpoints.
Research & Publications
Microfinance Digitalization
ICCIT 2025 (Accepted)
Designed inclusive digital financial services for low-literate populations, focusing on Human–Computer Interaction (HCI) principles and usability testing. Developed a mobile-first prototype that demonstrated improved accessibility for rural microfinance users.
PakhiderChobi
ICCIT 2023 (Accepted)
Created a dataset of 8,670 annotated images across 33 Bangladeshi bird species, enabling large-scale fine-grained classification. Achieved 95.3% mAP with YOLOv8, establishing a strong baseline for biodiversity monitoring in Bangladesh.
BanFish
ICEEICT 2024 (Accepted)
Developed the largest Bangladeshi fish image dataset, covering over 50 species, to support automated aquaculture monitoring. Applied Vision Transformers (ViT) for classification, achieving state-of-the-art accuracy compared to CNN-based baselines.
Bengali Document Layout Analysis
Detectron2
Conducted document segmentation on the BaDLAD dataset using Mask R-CNN in Detectron2. Improved layout parsing across diverse document types (books, newspapers, forms), addressing multilingual OCR challenges.
Projects
JobHub
A comprehensive job portal web application built with the MERN stack (MongoDB, Express, React, Node.js). Features dynamic job filtering, user authentication, and interactive dashboard statistics.
CP Mentor
Mobile application designed to guide beginners in competitive programming. Built with Android and Java, featuring Firebase Authentication and Firestore database for real-time user data.