SOFTWARE ENGINEER & ML RESEARCHER

MD RIFAT HOSSEN

Full-Stack Software Engineer

Full-stack developer and applied AI researcher specializing in scalable web and mobile systems, low-resource Bengali NLP, and explainable computer vision. Technical Lead at KREMS Technologies with 6 peer-reviewed papers published in Springer Nature (Q2 Journal) and IEEE.

KEY HIGHLIGHTS //

06

Published Papers

3.60

Graduated CGPA

LEAD

KREMS Technologies

834+

Active Transit Users

Download CV
Md Rifat Hossen
Available for Software & AI Engineering Roles
WORK & LEADERSHIP EXPERIENCE

01 // WORK EXPERIENCE

Software engineering, startup architecture, and technical team leadership in web and AI systems.

KREMS Technologies Logo

AI Engineer & Full-Stack Developer

Co-Founder & Technical Lead
KREMS Technologies•Remote / CUET ITBI Level-3•Startup
2025 - Present

Co-founded AI-powered software startup; architect full-stack applications, implement AI/ML pipelines, and direct technical strategy for an engineering team of 5. Secured office space at CUET IT Business Incubator (Level-3) to scale startup operations.

KEY DELIVERABLES & IMPACT //

→

Architected and delivered 10+ client software projects including HealthPort (AI healthcare), RChatbot (RAG chatbot), and official agency Next.js platform

→

Integrated machine learning pipelines and agentic automation workflows for client applications

→

Responsible for system scalability, database schema design, and technical roadmap execution

→

Collaborative technical leadership, mentoring, and code reviews across distributed team members

→

Secured office space at CUET IT Business Incubator (Level-3) to scale startup operations

Next.js 16TypeScriptReact 19Node.jsPythonAgentic AIFirebaseTailwind CSS v4MongoDB
W3

Full-Stack Web Developer

Industrial Attachment
W3 Eden•Chattogram, Bangladesh•Contract
3 Weeks Intensive

Completed intensive industrial training in modern production web development stack (React, Node.js, MongoDB, Firebase). Engineered CUET FoodExpress, an end-to-end campus cafeteria ordering platform with Stripe payments and administrative dispatch console.

KEY DELIVERABLES & IMPACT //

→

Engineered CUET FoodExpress end-to-end full-stack campus food ordering platform

→

Implemented Firebase authentication, JWT-secured Express endpoints, and Stripe card payment processing

→

Designed customer ordering UI and comprehensive kitchen admin management dashboard with sales analytics

React 18Node.jsExpress.jsMongoDBFirebaseStripeTailwind CSS
SOFTWARE & AI ENGINEERING

02 // FEATURED PROJECTS

Full-stack web applications, native mobile architectures, and applied machine learning models built for production.

MyCUETBus
🚌Mobile

MyCUETBus

Real-time campus transit tracking Android application with interactive Mapbox rendering and an MQTT-to-Firebase telemetry bridge streaming live GPS coordinates. Adopted by 834+ verified CUET students.

React NativeExpo RouterMapbox GLFirebase+2
CXR-Sentinel
🩻Web, ML/AI

CXR-Sentinel

AI-assisted chest X-ray analysis and clinical reporting prototype simulating a 5-stage radiological workflow with Gemini API, RAD-DINO ViT-B/14, and dual uncertainty gating (0.988 mean AUC).

Next.js 16TypeScriptPyTorchGemini API+2
CUET FoodExpress
🍕Web

CUET FoodExpress

Full-stack campus food ordering and cafeteria management platform featuring Firebase authentication, JWT-secured Express endpoints, Stripe card checkout, and an administrative dispatch dashboard.

React 18Node.jsExpress.jsMongoDB+3
Multi-Task Fruit Classification & Freshness XAI
🍎ML/AI

Multi-Task Fruit Classification & Freshness XAI

Unified multi-task ResNet152V2 model for simultaneous 9-class fruit categorization (99.64% acc) and freshness classification (97.27% acc, 0.998 ROC-AUC) with dual-head Grad-CAM XAI. Published in IEEE ICCIT 2025.

PythonTensorFlowResNet152V2Grad-CAM+2
KREMS Technologies Platform
🏢Web

KREMS Technologies Platform

Modern agency platform for AI automation engineering, RAG chatbot implementations, and full-stack software development with integrated EmailJS inquiry pipeline.

Next.js 16React 19TypeScriptTailwind CSS v4+1
Siyam IT Center
💻Web

Siyam IT Center

Production-grade multilingual IT agency website engineered with Next.js App Router, SSR/SSG, and zero-downtime CI/CD deployment on Vercel Edge Network.

Next.js 16React 19TypeScriptTailwind CSS+1
GitHub Repositories Archive →
PEER-REVIEWED SCIENTIFIC RESEARCH

03 // RESEARCH & PUBLICATIONS

Peer-reviewed academic research in deep learning defect classification, computer vision, and Bengali natural language processing.

SpringerSpringer Nature — SN Computer Science (Q2 Journal, Vol. 7, Article No. 446)
2026

BanglaSentNet: An Explainable Hybrid Deep Learning Framework for Multi-Aspect Sentiment Analysis with Cross-Domain Transfer Learning

Weighted F1 / Accuracy0.880 F1 / 85.0%

Dataset: 8,755 Real-World Bengali E-Commerce Reviews

Springer Q2 JournalBanglaBERTXAI / SHAPCross-Domain TransferBiLSTMFastText
SpringerSpringer ICDSAIA 2025 (CCIS Vol. 2682, pp. 283–300)
2025

BanglaSentNet: A Hybrid Deep Learning Framework for Multi-Aspect Sentiment Analysis in Bangla E-Commerce Reviews

F1 / Accuracy Score0.88 F1 / 85.0%

Dataset: 8,755 Verified Bengali Reviews

Springer CCISBanglaBERTGloVe 2.5BAspect SentimentDynamic Ensemble
IEEEIEEE ICCIT 2025 (28th International Conference on Computer and Information Technology)
2025

A Multi Task Deep Learning Model for Fruit Detection and Freshness Classification with GradCAM Explainability

Variety / Freshness Acc99.64% / 97.27%

Dataset: 30,357 Multi-Category Fruit Images

ResNet152V2Grad-CAM XAIMulti-Task LearningROC-AUC 0.998Hugging Face Demo
TECHNICAL STACK & PROBLEM SOLVING

04 // SKILLS & PROBLEM SOLVING

Core engineering stack, system design tooling, and competitive programming track record across major platforms.

Languages

PythonJavaScriptTypeScriptJavaC++CSQLBashSolidityHTML5/CSS3

Frontend & Mobile

React 19Next.js 16React NativeExpo RouterTailwind CSS v4NativeWindVite

Backend & Systems

Node.jsExpress.jsJava Spring BootRESTful APIsJWT AuthStripe APIFirebase RTDB

AI, ML & Research

PyTorchTensorFlowscikit-learnBanglaBERTResNetGrad-CAM XAISHAPLLM APIs (Gemini)PandasNumPy

Databases & Cloud

MongoDBMySQLFirestoreFirebase HostingVercelRenderRailwayDocker (basics)Git & GitHub
ACADEMIC & PROFESSIONAL CREDENTIALS

05 // EDUCATION & CREDENTIALS

Academic degree in Computer Science & Engineering at CUET alongside verified professional specializations and leadership tenures.

B.Sc. in Computer Science & Engineering

CGPA: 3.60 / 4.00

Chittagong University of Engineering & Technology (CUET)

Graduated in CSE. Co-authored 6 peer-reviewed research papers in deep learning and NLP published in Springer Nature (Q2 Journal) and IEEE. Led campus initiatives, hackathons, and software engineering projects.

2022 – 2026

Chattogram, Bangladesh

Higher Secondary Certificate (HSC) — Science

GPA: 5.00 / 5.00

Dhaka College

Science Group. Core coursework in Physics, Chemistry, Higher Mathematics, and Biology, along with General curriculum subjects including ICT, English, and Bangla. Graduated with GPA 5.00 / 5.00.

2018 – 2020

Dhaka, Bangladesh

Secondary School Certificate (SSC) — Science

GPA: 5.00 / 5.00

Samsul Haque Khan School and College

Science Group. Core coursework in Physics, Chemistry, Higher Mathematics, Biology, and General Science, with general curriculum subjects in Bangla, English, and Social Studies. Graduated with GPA 5.00 / 5.00.

2013 – 2018

Dhaka, Bangladesh

CREATIVE MEDIA & VISUAL ARTS

06 // CREATIVE STUDIO & BEYOND CODE

Cinematic travel filmmaking, visual branding, technical writing, and strategic pursuits.

Rifat HossaiN.

@RifatHossaiNBro

Travel Filmmaking & Cinematic Visuals

Documenting raw landscapes, mountain trails, and cultural journeys across Bangladesh. Combining cinematic drone perspectives, dynamic sound design, and custom color grading to tell immersive travel stories.

Travel DocumentariesCinematic FilmsWilderness ExpeditionsDrone Cinematography

CINEMATOGRAPHY & POST-PRODUCTION WORKFLOW //

Cinematic Video Editing

DaVinci Resolve, Adobe Premiere Pro, CapCut

Camera & Drone Composition

4K Composition, Framing, Gimbal Tracking

Color Grading & LUTs

Log Correction, Film Look, Ambient Tonal Balance

Sound Design & Atmosphere

Foley, Environmental Audio, Dynamic Pacing

GET IN TOUCH & CONNECT

07 // CONTACT & COLLABORATION

Available for software engineering positions, peer-reviewed academic research collaborations, and technical consulting.

DIRECT TELEMETRY // EMAIL

rifat8851@gmail.com

Click icon to copy

Chattogram / Dhaka, Bangladesh
Software Engineering & AI Systems

NETWORKS // CODE

Dispatch a Direct Message