Arnab

D.M. Rafiun Bin Masud
dmrafiun@gmail [DOT] com

I am working as an AI Engineer at SM Technology, Dhaka, Bangladesh.
AI Engineer with hands-on experience turning AI concepts into production-ready applications. Skilled in Python, TensorFlow, PyTorch, Scikit-learn, OpenCV, and FastAPI, with experience across the machine learning lifecycle, including data preprocessing, model development, API integration, and Dockerized deployment. Particularly interested in Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), AI Agents, Computer Vision, and MLOps, with a strong focus on combining practical engineering with applied AI research.

CV / GitHub / LinkedIn / Kaggle / ResearchGate

Education

B.Sc in Computer Science & Engineering (CSE) - 2024
East West University (EWU)

Work Experience

AI Engineer | SM Technology, Dhaka, Bangladesh
March 2025 – Present

Design and development of scalable AI backend systems using Python and FastAPI to power intelligent web and mobile applications.
Core focus areas: Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), Natural Language Processing (NLP), Computer Vision, and AI-driven automation.
Emphasis on building production-ready, robust, and maintainable AI solutions for real-world deployment.

Research Assistant | East West University, Dhaka, Bangladesh
Jan 2026 – Present

Completed a four-member capstone project under the supervision of Dr. Md. Sawkat Ali, Associate Professor.
Led project planning, task allocation, stakeholder communication, and progress reporting to ensure timely delivery.

IT Support Engineer | Taltola Dot Net, Dhaka, Bangladesh
March 2021 – January 2022

Managed end-to-end network operations, including configuration, installation, maintenance, and upgrades.
Proactively monitored IT systems to detect and resolve performance bottlenecks before they impacted users.

Technical Skills

Programming Languages:	Python, C++, C, Java, PHP, SQL
AI/ML & Generative AI:	TensorFlow, PyTorch, Keras, Scikit-learn, Transformers, Hugging Face, NLTK, NLP, Large Language Models (LLMs), RAG, LangGraph, Explainable AI (XAI), Transfer Learning
Computer Vision & Speech:	OpenCV, Image Classification, Object Detection, Data Augmentation, Adversarial Training, Speech Recognition, Audio Processing
Backend & APIs:	FastAPI, REST APIs, Async Processing, Microservices, Postman
Databases & Vector Stores:	MySQL, Oracle SQL, Vector Databases
Cloud & DevOps:	Docker, Git, GitHub, GitHub Actions, CI/CD, nginx, AWS, GCP, VPS Deployment
Data Science & Analytics:	Pandas, NumPy, Matplotlib, Seaborn, EDA, Feature Extraction, Statistical Analysis
Web Technologies:	React, HTML, CSS

Projects

[1]	Apple Plum Growth Stage Classification: Curated a 701-image agricultural dataset and engineered preprocessing pipelines and CNN architectures for growth-stage classification and bounding-box detection. (TensorFlow, Keras, OpenCV, Streamlit)
[2]	MANIFEX AI Paragraph Evaluator: Built an NLP engine that evaluates sentence structure, grammar, vocabulary richness, and overall writing quality with automated scoring. (NLP, LLM, Docker)
[3]	Piper TTS FastAPI Inference Server: Built a low-latency, containerized neural text-to-speech server delivering real-time voice synthesis via high-throughput REST endpoints. [GitHub] (FastAPI, Docker, REST API)
[4]	Movie Recommendation Engine: Built a dual-engine recommender combining content-based cosine similarity and collaborative-filtering matrix factorization on large user-rating matrices. (Machine Learning, Scikit-learn, Python)
[5]	DFT Image Prep & Enhancement: Performs super-resolution, background removal, and automatic trimming using advanced image processing algorithms. [GitHub]
[6]	Bilingual Bengali Medicine Recommendation Chatbot: Integrated Bengali language support for medicine information and designed/trained a chatbot using large language models to offer accurate medical advice. (Python, Streamlit, LLMs)
[7]	MedGemma-Based Smart Healthcare Platform: Wound analysis, surgical outcome simulation, and injectable medicines guidance with NHS Injectable Medicines.
[8]	Qur’an Recitation Pronunciation Error Detection: Multilingual AI framework using speech recognition and language models. Sunnahshelf

Certifications

Accenture North America - Data Analytics and Visualization Job Simulation - 2024 | Forage

Volunteer Experience

Event Coordinator | IEEE Computer Society Student Branch, EWU
Jan 2022 – Aug 2023
Coordinated and organized technical events, workshops, and student activities to ensure smooth execution.

Associate Video Editor | EWU Robotics Club
Collaborated with team members to create visually appealing content that showcased robotics projects and achievements.

Research Papers

[1]	Deep Learning and Non-Contrast CT for Widespread Pancreatic Cancer Detection D.M. Rafiun Bin Masud, Wasif International Conference on Innovations in Data Analytics (Springer ICIDA), 2022 🏆 BEST PAPER AWARD [DOI: 10.1007/978-981-96-6303-3_38] \| [Achievement Link] Investigated applying deep learning models to non-contrast CT scans for early detection of pancreatic cancer. Designed and trained CNNs to accurately identify cancerous tissues.
[2]	A Machine Learning Approach to Evaluate the Classification of Slow Growth Stages in the Ball Shundari Apple Plum Image Dataset D.M. Rafiun Bin Masud Academic Project, 2022 Developed a machine learning model to classify the growth stages of the Ball Shundari Apple Plum based on image datasets. Applied feature extraction techniques and optimized classification algorithms for accuracy.
[3]	A Deep Learning Approach to Classify Slow Growth Stages in the Ball Shundari Apple Plum D.M. Rafiun Bin Masud Research Project, 2021 Implemented a deep learning framework to analyze and classify slow growth stages in apple plum samples. Utilized neural networks to model intricate growth patterns and detect stage-specific features.
[4]	ShrimpXNet: A Transfer Learning Framework for Shrimp Disease Classification with Augmented Regularization, Adversarial Training, and Explainable AI D.M. Rafiun Bin Masud IEEE Conference Publication, 2025 [DOI: 10.1109/3ict68299.2025.11442114] \| [Paper] Developed a transfer learning-based framework for automated shrimp disease classification. Integrated augmented regularization, adversarial training, and explainable AI techniques to improve model robustness, classification accuracy, and interpretability.
[5]	SFiNNet: Leveraging Transfer Learning and Explainable AI for Automated Small Fish Classification Enhanced via Adversarial Training and Augmented Regularization D.M. Rafiun Bin Masud IEEE Conference Publication, 2025 [DOI: 10.1109/ICCIT68739.2025.11491764] \| [Paper] Proposed an intelligent fish classification framework utilizing transfer learning, explainable AI, adversarial training, and augmented regularization. Enhanced classification performance while providing transparent and reliable predictions for small fish species identification.
[6]	Emerging Trends in Data Leakage Detection and Prevention: A Comprehensive Review Nibbritta Niloy Sarker, Amrita, D.M. Rafiun Bin Masud, Avinash Kumar Sharma International Conference on Advances in IoT and Security with AI (ICAISA), Springer LNNS, 2026 [DOI: 10.1007/978-3-032-10386-4_38] \| [Springer Link] A comprehensive review exploring recent developments in data leakage detection and prevention, evaluating machine learning, anomaly detection, deep learning, deception tactics, and blockchain technologies for secure data handling.