M.Sc.
Marquette University
Master's in Data Science · 2026 — 2027
Research Assistant (mHealth)
Mazhar Hossain
Primary stack
- Python
- PostgreSQL
- Airflow
- PyTorch
- Docker
Impact
- 500+ data sources integrated
- 95% faster debugging via automation
- 60% less effort onboarding new sources
- 35% faster incident response time
About the work
Python backend and data engineer with 6+ years across data pipelines, ML systems and production APIs. Focused on data quality, observability and scalable ingestion frameworks. Research Assistant at Marquette University (mHealth), currently developing ML-enabled mHealth data collection platforms.
Data Integrity
Monitoring
Prediction
Automation
Top Skills
- Python & Backend Systems
- ETL & Automation
- Data Quality & Observability
0
Years in data & ML engineering
0
Data sources integrated
0
Stack Overflow reputation
SO ↗0
Google Scholar citations
Scholar ↗Work history
Marquette University
Research Assistant
mHealth research, applied ML & biomarker extraction from face/fingertip video.
mHealth research, applied ML & data processing for health-focused studies.
- Research on mHealth applications and applied ML.
- Designing & developing a suicide prevention survey system for early risk detection & behavioral assessment.
- Building multimodal patient monitoring frameworks for mental & physiological health tracking.
- Researching neuro-signal correlations with physical health indicators.
- Designing computer vision methods for contactless health monitoring via face and fingertip video signals.
- Extracting remote biomarkers including immunization & white blood cell (WBC) indicators.
Cefalo
Software Engineer → Senior
500+ source pipelines, 95% faster debugging via automation, ingestion framework that cut new-source effort 60%.
Data platform reliability, 500+ source pipelines & 95% faster debugging via automation.
- Built energy market forecasting data products and RESTful backends.
- Built RESTful services for urgent market data ingestion.
- Managed pipelines from 500+ sources with high integrity and availability.
- Automated verification processes, reducing debugging time by 95%.
- Redesigned ingestion framework, cutting new-source effort by 60%.
- Implemented monitoring for critical sources, reducing response time by 35%.
- Contributed to data platform reliability and automation.
Upwork
Python Developer (Freelance)
ETL delivery across diverse sources, system stabilization, 100% job success rate.
ETL delivery, system stabilization & 100% job success rate.
- Delivered ETL pipelines across diverse data sources.
- Stabilized systems by fixing and optimizing existing scripts.
- Produced clear documentation for workflows and installations.
- Maintained 100% job success rate.
BJIT
AI Engineer
Drone object detection at 74% accuracy, NLP chatbots +43%, recommender systems.
Computer vision (74% detection accuracy), NLP chatbots (+43% accuracy) & recommender systems.
- Improved dataset quality by 30% through augmentation and labeling.
- Delivered drone small-object detection with 74% accuracy.
- Improved chatbot response accuracy by 43% with model training and API integration.
- Built recommender system using collaborative and content-based filtering.
Education & foundations
B.Sc.
Khulna University of Engineering & Technology
B.Sc. in ECE · 2014 — 2018
Electronics & Communication Engineering
Publications
Employing PCA and t-statistical approach for feature extraction and classification of emotion from multichannel EEG signal
Egyptian Informatics Journal · Nov 2, 2019
- Hybrid PCA + t-statistics features for EEG emotion recognition on SEED.
- Evaluated with SVM, ANN, LDA, k-NN across subject-dependent and independent setups.
- Best accuracy: 86.57% (ANN, subject-dependent).
Toolkit & stack
Python
ETL, automation & production APIs.
PostgreSQL
Data modeling, optimization & reporting.
Django
REST APIs and data-driven applications.
Docker
Containerized deployments and local dev flows.
Airflow
ETL
Web Scraping
PyTorch
Scikit-learn
RASA
OpenCV
GitHub Actions
Prometheus
Grafana
Selected projects
Chatbot Platform
Intent- and entity-based chatbot system with dataset tooling, multi-profile support & Flask APIs.
Small Object Detection
Trained YOLOv3 and Faster R-CNN pipelines for drones, birds, and underwater fish with performance reporting.
Django Web App
Web app for server lookup by IP and content recommendations.
Scraping Solutions
Data extraction tools using multiple scraping frameworks with production-ready data cleaning.
Certifications & credentials
OWASP Top 10:2021 in Python
SQL for Data Science
Python Specialization
Deep Learning Specialization
Machine Learning
Let's build something
Open to opportunities in data engineering, analytics & ML systems. Collaboration, consulting, or a quick chat about a gnarly pipeline: pick whichever door you prefer.