Resume

I’m a data scientist motivated by turning research into real-world impact. My work bridges NLP, computer vision, and health analytics, creating systems that feel intuitive and useful, from emotion-aware interfaces to contactless cardiovascular monitoring and AI-driven tools for education and telehealth.

I focus on more than models and metrics. I care about what they enable: faster responses in elder care, stronger engagement in therapy and human–robot interaction, and accessible learning for the next generation of AI builders.

This page reflects both the solutions I have built such as LLM-based QA agents, NER pipelines, facial emotion recognition, and gesture detection systems, and the challenges I am still exploring to push human-centered AI forward.

On this page

Latest work experience

In my most recent work, I contributed to real-time machine learning systems in telehealth and human–computer interaction. This included developing LLM-based agents with NER, sentiment analysis, and QA, a cardiovascular detection agent, and a contactless heart rate monitoring system for remote elderly care. I also improved facial emotion and gesture recognition by integrating landmark features with ML, raising accuracy by over 15%. Across these projects, I focused on combining deep learning, computer vision, WebRTC, and privacy-preserving design to deliver accessible, human-centered AI solutions.

“He builds intelligent systems that see what people often miss and does it with quiet precision, not showmanship.”

Lead Clinical Technology Advisor, Digital Mental Health Platform

EHR – AI-Enabled Electronic Health Records Platform

Duration: May 2025 – Aug 2025

Designed and developed a secure, AI-enabled EHR platform for a non-profit healthcare institute serving low-income and remote communities. The system streamlined patient registration, medical history tracking, and health risk detection, while ensuring privacy through encryption and role-based access control. Optimized for low-resource environments, the platform empowered doctors, nurses, and volunteers to deliver faster, more accurate care and maintain continuity of service.

5 Most used Software skills

Speech-to-Text Processing

Named Entity Recognition (NER)

API & Microservices Integration

Docker & API Deployment

Containerized Deployment

FER – Facial Expression Recognition for Teletherapy

Duration: Jan 2022 – Mar 2024

Led the development of a real-time FER system to assist psychologists during remote consultations. Combined facial landmark detection, TensorFlow emotion models, and WebRTC for secure, low-latency deployment. The platform helped clinicians detect emotional cues, enriching patient documentation while preserving privacy through opt-in-only processing.

5 Most used Software skills

Real-Time Machine Learning

Facial Landmark Detection

WebRTC & WebSocket Integration

Docker & API Deployment

Privacy-Preserving Design

BPM – Contactless Heart Rate Monitoring

Duration: Sep 2022 – Apr 2024

Created a real-time, non-contact heart rate monitoring system using facial motion analysis. Developed for elder care, the system combined WebRTC video streaming with facial tracking and micro-signal extraction, offering caregivers a wearable-free alternative with live WebSocket alerts and high reliability across different lighting and video conditions.

5 Most used Software skills

Signal Processing from Facial Landmarks

Real-Time Video Streaming

WebRTC & WebSocket Pipelines

OpenCV Feature Extraction

Non-Invasive Health Monitoring

HPO – AI Education Platform

Duration: Feb 2023 – Feb 2025

Built an interactive AI learning platform allowing students to train, tune, and visualize deep learning models in real time — without writing code. Designed and deployed using TensorFlow, Streamlit, and Kubernetes, the tool simplified ML concepts and supported hands-on education through an intuitive, responsive, classroom-ready interface.

5 Most used Software skills

Streamlit UI Development

TensorFlow/Keras Model Training

Hyperparameter Tuning & Visualization

Docker & Kubernetes Deployment

Interactive Educational Design

Qualification

With expertise in real-time AI deployment, machine learning, human-centered system design, and full-stack development, I specialize in building intelligent systems that serve real-world needs in healthcare, education, and human-computer interaction.

Certificate

I’ve earned certifications from Udacity, Google Cloud, and DataCamp, highlighting my commitment to applied machine learning, large language models, and deployment tooling. These credentials reflect hands-on expertise in natural language processing, deep learning, and real-world AI integration with ethical and scalable practices.

Udacity Certificates:

DataCamp Stack (Merged PDF):

Competence

Designing and deploying ML pipelines

Real-time inference and signal processing

WebRTC communication

Interactive AI tool development for education

Human-centered system design and ethics in AI

Facial emotion recognition and biometrics

Docker/Kubernetes MLOps workflows

Privacy-aware data streaming

RESTful API and WebSocket integration

Expertise

Machine Learning, Real-Time Systems, AI Ethics, Emotion Recognition, Data Visualization, Educational AI Tools, Streamlit Development, Hyperparameter Optimization, Full-Stack ML Deployment, MLOps (Docker/Kubernetes), WebRTC + WebSocket Streaming, Consent-Driven Design, Health Tech, DevOps for ML, Human-AI Interaction

Most Used Skills

Real-Time ML Inference, WebRTC Pipelines
0%
Streamlit, Hyperparameter Tuning, Interactive UI
0%
TensorFlow, Docker, Deployment Automation
0%

Software skills

I work across the full machine learning stack with tools that support real-time performance, modular design, and scalable deployment. From data pipelines to interactive UIs and containerized services, I rely on frameworks like TensorFlow, Streamlit, and OpenCV to move fast without compromising quality. My workflow is powered by Docker, Kubernetes, Git, and modern API architectures — all orchestrated to deliver responsive, production-ready systems.

Most Used Skills

Python, TensorFlow, Scikit-learn
0%
OpenCV, MediaPipe
0%
Streamlit, FastAPI, WebSocket, REST
0%
Docker, Kubernetes, Nginx
0%
SQL, Git
0%
Google Colab, Jupyter, VSCode
0%
CI/CD & Cloud-based deployment
0%
Linux CLI, Shell scripting
0%

Education

I completed my academic training at Iran University of Science & Technology and the University of Tehran, where I developed a strong technical foundation grounded in systems thinking, practical problem-solving, and the ability to work across disciplines and technologies.

Building on this foundation, I continued my education at the University of Coimbra in Portugal, where I successfully defended my PhD proposal. This phase allowed me to deepen my applied expertise and focus on building intelligent systems that address real-world needs in scalable and impactful ways.

Resume Preview

I studied “Integrated Design” at the Cologne International School of Design in Germany. The Bachelor Degree Program there integrates different design and scientific approaches from several design areas and design-related disciplines.

The study programme is characterized by an integrated, intercultural and interdisciplinary environment in which theory and practice-oriented design processes are complemented by a range of experimental and research-oriented elements.

Let’s Connect

Are you working on a project that bridges machine learning, real-time systems, or digital health?
Whether you’re building something innovative, looking for a technical collaborator, or just want to exchange ideas, I’d love to hear from you.
Feel free to reach out for a chat about projects, collaborations, or research.
Email me at: contact@hosamzolfonoon.pt
Let’s build technology that truly makes a difference.

HOSAM ZOLFNOON

ML Engineer

contact@hosamzolfonoon.pt 🔗
www.hosamzolfonoon.pt 🔗

Feel free to contact me

on social media.

Introduction

Portfolio

Resume

Copyright © 2025 | Hosam Zolfonoon