Md Rana Hossain

Computer scientist · Detroit, Michigan, USA

Trustworthy health-AI research
and full-stack AI engineering

Peer-reviewed methods for leakage-aware, calibrated clinical-risk modeling—and production RAG, agentic systems, and MCP-backed platforms. Interested in a funded Ph.D. in CS / health informatics.

Academia · Health AI research Ph.D. · Funded research Industry · Full-stack AI
Research agenda Publications Experience Scholar ORCID

M.S. Computer Science (GPA 3.93) · Senior Full Stack Engineer, Volkswagen Group of America
mdrana.hossain@miu.edu · ranasl62@gmail.com

VW GoASenior Full Stack · RAG / agents
LangGraph · MCPProduction AI systems
IEEE / SpringerPeer-reviewed publications
First authorHealth imaging research
4 WoSClarivate-verified reviews
Open softwareEHR Risk Framework

About

Publishing trustworthy health-AI methods and shipping production AI systems

Md Rana Hossain · Detroit, Michigan · Senior Full Stack Engineer · Independent researcher

Portrait of Md Rana Hossain
Md Rana Hossain · Detroit, Michigan

I work across academic research and industrial engineering: leakage-aware, calibrated, explainable methods for EHR risk and medical imaging, and grounded RAG and agent platforms in production.

As a Senior Full Stack Engineer at Volkswagen Group of America, I deliver LangChain/LangGraph retrieval systems, multi-step agents, MCP tool wiring, LangSmith tracing, chat assistants, and task automation on Spring Boot and Angular platforms.

Independently, I publish peer-reviewed health-AI work, maintain the open EHR Risk Framework, and serve as a peer reviewer—including Web of Science–verified journal reviews. That scholarly record is separate from employer product ownership.

Healthcare methods are research and education only—not a medical device and not bedside deployment.

Research

Health AI methods & open software

Agenda: honest temporal evaluation of EHR and imaging models—index-time contracts, patient-level splits, calibration, and SHAP—released as papers and inspectable tools.

EHR Risk Framework

Creator and maintainer. Open workbench for longitudinal EHR-style binary risk: index/horizon cohorts, multi-window lookback, patient-level splits, leakage audits, isotonic calibration (Brier, ECE), TreeSHAP, fairness slices, JSON manifests, FastAPI + Angular UI.

MIT License · synthetic verification only · not clinical performance claims

GitHub Zenodo Documentation Demo

Methods book

Artificial Intelligence in Healthcare: Early Disease Detection and Clinical Risk Prediction from Electronic Health Records — lead author, First Edition 2026.

Harithpatra Prokashon, Dhaka · ISBN 978-984-29284-1-3

Doctoral research interests

I am interested in a funded Ph.D. focused on trustworthy clinical prediction from real-world health data. Proposed dissertation anchor:

How much do leakage and miscalibration inflate reported performance in EHR risk studies—and can automated index-time audits plus calibration benchmarks reduce that inflation on credentialed cohorts?

  • Temporal integrity / leakage audits for EHR risk models
  • Probability calibration and reliability under shift
  • Open, reproducible research software for multi-site evaluation

Faculty: mdrana.hossain@miu.edu · Scholar and ORCID in the header.

Scholarship

Peer-reviewed publications

Selected work first. Author roles match publisher lists. Full indexed record: Google Scholar.

Selected

  1. Co-author Asad, M., Khan, M., Hossain, M.R., et al. CRuSE-Heart: Explainable AI for Risk-Controlled Selective Ensembles in Heart-Attack Screening. Discover Artificial Intelligence (Springer Nature), 2026. doi:10.1007/s44163-026-01378-x
  2. First author Hossain, M.R. et al. Hybrid Deep Learning Framework for Knee Osteoporosis Diagnosis Using CNNs and Fine-Tuned Multimodal Large Language Models. ICSADL, IEEE, 2026. doi:10.1109/ICSADL67539.2026.11451868
  3. Lead author Hossain, M.R. Artificial Intelligence in Healthcare: Early Disease Detection and Clinical Risk Prediction from Electronic Health Records. First Edition, Harithpatra Prokashon, 2026. ISBN 978-984-29284-1-3.
  4. Co-author Early Diagnosis of Diabetic Retinopathy Through AI-Based Optimized Ensemble Classification. ICARC, IEEE, 2026. doi:10.1109/ICARC68737.2026.11453518
  5. Co-author Optimizing Transfer Learning: A Deep Learning Approach for High-Accuracy Classification of Alzheimer’s Disease Stages. AISEI, IEEE, 2026. doi:10.1109/AISEI68628.2026.11572889

Additional peer-reviewed

  1. Co-author Advancing Public Safety with Real-Time Life Jacket Detection and Demographic Profiling Using YOLOv8 and Age Classification. EAI Endorsed Transactions on AI and Robotics, 2025. doi:10.4108/airo.9785
  2. Co-author Enhanced Colorectal Cancer Image Classification Using a Custom Deep Convolutional Neural Network. ICSADL, IEEE, 2026. doi:10.1109/ICSADL67539.2026.11452002
  3. Co-author Hybrid Vision Transformer and CNN Architectures for Multi-Class Classification in Gastrointestinal Endoscopy. AISEI, IEEE, 2026. doi:10.1109/AISEI68628.2026.11572866
  4. Co-author Revolutionizing Drug Discovery: A Systematic Review of AI and Machine Learning Application. ICSSAS, IEEE, 2025. doi:10.1109/ICSSAS66150.2025.11081368
  5. Co-author Automated Jute Leaf Disease Detection Using CNNs. ICSADL, IEEE, 2026. doi:10.1109/ICSADL67539.2026.11451802
  6. Co-author Bangladeshi License Plate Detection and Recognition Using YOLO Variants and Enhanced OCR. RTIP2R, Marrakech, 2025.
  7. Co-author Transparency in Fabric Defect Detection by Leveraging Explainable AI. RTIP2R, Marrakech, 2025.
  8. Co-author Image Recognition for Early Diagnosis and Prognosis of Alzheimer’s Disease Using Multimodal Brain Imaging. ICEFronT 2026. Accepted (DOI forthcoming).

Under review

  1. FWC-TabXAI++: Feature-Weighted Calibrated Tabular Explainable AI for Heart Disease Prediction. Discover Artificial Intelligence (Springer). Co-author.
  2. Enhanced Brain Tumor Diagnosis Through Multi-Modal Feature Analysis and Hyperparameter-Tuned XGBoost. Expert Systems With Applications (Elsevier). Co-author.
  3. Integrating Custom CNN and Explainable AI for Evaluating GNNs and CNNs in Brain Tumor MRI Classification. Healthcare Analytics (Elsevier). Co-author.

Industry

Professional experience

U.S. production AI and full-stack delivery, with prior platform and FinTech leadership.

Senior Full Stack Engineer

Volkswagen Group of America · Auburn Hills, MI · Apr 2024 – present · Hybrid

  • Production LangChain / LangGraph RAG for repair-manual search with cited answers (Solr chunking and indexing) on dealer and technician platforms.
  • Multi-step agentic workflows (Infomedia, customer-feedback analysis); MCP tool wiring; LangSmith tracing and evaluation.
  • Internal chat assistants over grounded technical knowledge; task automation to reduce manual search and handoffs.
  • Spring Boot APIs and Angular 18 UI; led Angular 4→18 with NgRx; mentor engineers.
  • ELSA2Go / Infomedia on AWS; Connected Car, myVW+, SiriusXM, CARFAX integrations.
  • CI/CD (Docker, GitHub Actions, Jenkins); JUnit, Mockito, SonarQube.

Java · Spring Boot · Angular · NgRx · Python · Solr · LangChain · LangGraph · MCP · LangSmith · Docker · AWS

Software Engineer & Director

The Tech Nerds Limited · Chattogram, Bangladesh · Oct 2021 – Jan 2023

  • Director-level delivery plus day-to-day engineering: monolith → microservices (Spring Boot, NestJS, React), JWT / Spring Security.
  • PostgreSQL/MySQL performance, dynamic form engine, CI/CD on Jenkins and AWS; Kafka, GraphQL, Kubernetes, Terraform.

Software Specialist, previously Associate

SSL Wireless · Dhaka, Bangladesh · FinTech / payments · May 2018 – Sep 2021

  • Specialist (2021): microservices split of high-traffic recharge (Spring Boot, Node.js, AWS); QR payments, Kafka, ELK, eKYC banking integrations.
  • Associate (2018–2020): Angular / Spring Boot payments and banking; legacy modernization.
  • Employee of the Month, December 2020.

Also independent contractor / freelance work with treffio.com (event platforms), Last Call Media (time-tracking and Next.js/NestJS/AWS), and VioResume (AI resume and job-matching product).

Engineering

Production AI stack

Summary of systems used in professional delivery—detail lives under Experience.

RAG & agents: LangChain / LangGraph, grounded retrieval with citations, MCP tool wiring, LangSmith tracing, Solr indexing and vector search where needed.

Platforms: Java 21 / Spring Boot · Angular (4→18, NgRx) · TypeScript · Python · FastAPI · React · MySQL / PostgreSQL / MongoDB · Kafka · GraphQL · AWS · Docker · CI/CD.

Service

Peer review

Eight completed independent reviews. Four journal reviews are Clarivate / Web of Science verified (Publons). Profile: WoS reviewer record.

Journals (Clarivate verified)

  • Biomedical Physics & Engineering Express (IOP, ISSN 2057-1976) — 3 · Vovoxjz4, fYL4oLXU, TYwFtCpm
  • PLOS ONE (ISSN 1932-6203) — 1 · k2QCAslj (journal-verified)

Other journals & conferences

  • International Journal of Public Health Practice and Promotion — 1
  • Health Information Science (HIS 2026) — 1
  • aiHealth 2026 — 1
  • ICISSC 2026 — 1

Training: Nature Masterclasses Focus on Peer Review; Elsevier Researcher Academy modules (2026). ResearcherID: PUC-9517-2026.

Credentials

Education & professional standing

Education

M.S., Computer Science
Maharishi International University, Fairfield, IA
Feb 2023 – Oct 2025 · GPA 3.93/4.00

B.S., Computer Science and Engineering
Southeast University, Dhaka
2015 – 2019 · GPA 3.97/4.00 · First in class (batch topper)

Certifications

  • AWS Certified Solutions Architect – Associate (Credly)
  • Johns Hopkins / Coursera: Social and Technical Context of Health Informatics
  • IBM / Coursera: Vector DBs for RAG; Build RAG Applications; Build AI Agents using MCP
  • Hugging Face: AI Agents Fundamentals

Memberships

  • IEEE Professional Member
  • American Medical Informatics Association (AMIA)
  • HIMSS Digital Member

Selected awards

  • First in class (batch topper), B.S. CSE, Southeast University
  • Employee of the Month, SSL Wireless (Dec 2020)
  • ACM ICPC Dhaka Site, Certificate of Achievement (2018)
  • National Collegiate Programming Contest, MBSTU (2017)

Contact

Get in touch