📍 Dhaka, Bangladesh  ·  📫 marzun.enamul@nesia-inc.com

I'm a software and AI engineer at Nesia Inc. and a researcher working at the intersection of computer vision, retrieval-augmented generation, and agent-guided systems. I recently completed my B.Sc. in Computer Science and Engineering at the Military Institute of Science and Technology (MIST), Dhaka, where I co-authored conference papers in robotics, low-light image restoration, and medical image analysis.

• Computer Vision - image restoration, segmentation, object detection, and applied medical imaging
• Retrieval-Augmented Generation (RAG) - grounded LLM systems for high-stakes domains
• Agent-guided Systems - LLM-driven planning and orchestration for real-world workflows
• Adjacent interests - Quantum Computing · Functional Programming · Distributed Systems

Languages TypeScript · JavaScript · Java · Python · C / C++

ML / CV PyTorch · OpenCV · YOLO · U-Net · Image segmentation · Object detection · Jupyter

Frontend React · Next.js · Remix · Tailwind CSS

Backend Node.js · Hono · Effect-TS · Spring · Laravel

Cloud & Data GCP · Cloudflare Workers · PostgreSQL · MySQL · Oracle · Drizzle ORM · Prisma

• A Hybrid Residual U-Net Architecture for Enhanced Low-Light Image Restoration - Conference paper, July 2025. Deep learning architecture (hybrid residual U-Net) for image restoration in poor-lighting conditions.
• Design and Development of an RC Ornithopter for Remote Surveillance - Conference paper, July 2025. Bio-inspired flapping-wing UAV for agile low-speed surveillance flight.
• An Automated Framework for Segmenting and Identifying Thyroid Nodules in Ultrasound Images - Conference paper, December 2024. Computer-vision pipeline for computer-aided diagnosis of thyroid cancer from ultrasound.

📑 Full list, citations, and PDFs: Google Scholar · ResearchGate

B.Sc. in Computer Science and Engineering Military Institute of Science and Technology (MIST), Dhaka, Bangladesh - 2021 – 2025

• Software and AI Engineer, Nesia Inc. - AI-based system development.
• Software Engineer (Automation), MarginEdge (Dhaka team) - engineering for the U.S. restaurant management platform.

• Thyroid-Nodule-Detection - Automated segmentation and classification of thyroid nodules in ultrasound (companion code to the December 2024 paper).
• Fire_Detection_Notification_Real_time - Real-time fire detection on ESP32 video streams using YOLOv5.
• Deep-Cluster - Deep clustering experiments for unsupervised representation learning.
• QNN - Quantum neural network experiments.

See all research repos →

• fsd-skill - Feature-Sliced Design architecture guardian skill for React, Next.js, Remix, and React Router.
• goalFlow-agent - LLM-assisted goal and workload management agent (Hono · Effect-TS · React Router · Postgres).
• exppance-tracker - Full-stack personal-finance tracker on Cloudflare Workers (Hono · Drizzle · React).
• OAuth-2.0_TikTok_Shop - TikTok Shop OAuth 2.0 integration in TypeScript with Effect-TS layers.

See all engineering repos →

Selected projects from my B.Sc. at MIST.

• Bikreta_client / Bikreta_server - Bulk-purchasing platform with admin sales analytics (MERN · MongoDB).
• Transit-Pro - Combined bus management solution with secure backend (Laravel · Oracle).
• Article_Publish_Systemm - Collaborative platform for writers with user profiles (Java).
• One-Connect - Online verification and donation platform (Flutter · Firebase · Firestore).

See all undergrad repos →

• Effect-TS - Notes and experiments with Effect-TS (functional patterns, schema, error handling).
• Database-Essentials - Schema-first PostgreSQL setup (Atlas migrations · multi-tenant analytics).
• Data-Science-Pre-requisites - Beginner-friendly roadmap of math/stats/tooling before Python for DS.

See all learning notes →