Think in systems
Understand the complete problem and workflow before optimizing a single component.
I turn complex data into practical AI and reliable software.
AI / Machine Learning Engineer · Data Scientist · Researcher
I'm Amir, an AI / Machine Learning Engineer, Data Scientist and PhD candidate in Neuroscience. I combine statistical reasoning, model development and software engineering to solve problems across computer vision, medical imaging, clinical data and generative AI. My work connects research with practical systems—from validated models to APIs, retrieval platforms and deployment.
Machine Learning & AI✦Data Science & Statistical Modeling✦Software Engineering✦Applied AI & RAG✦Medical Imaging✦Computer Vision✦
01 · About
Engineering with depth, curiosity and context.
I work across AI engineering, machine learning, data science and computer vision, combining analytical thinking with practical system building.
My background spans electrical engineering, mathematics, statistics, software engineering and neuroscience. I enjoy understanding complex, open-ended problems, analyzing the data and underlying theory, experimenting with different approaches, and turning validated results into practical AI and software systems.
My experience includes medical-image segmentation, synthetic data, statistical and graph-based machine learning, NLP, reinforcement learning, scientific RAG platforms, robotics and backend development. During my PhD, I have further developed my skills in high-dimensional data analysis, statistical modeling, optimization and reproducible research.
Make it reproducible
Design experiments and software that can be tested, explained and repeated with confidence.
Protect the evidence
Preserve validation, provenance and context throughout data-driven workflows.
Build for change
Prefer maintainable architecture and clear interfaces over short-lived technical shortcuts.
02 · Experience & Projects
Research, engineering and projects in one connected path.
Research appointments, industry roles, collaborations and independent projects, with the methods and skills used in each.
2023 – PresentResearch
01
Research
Machine learning for post-stroke function and brain networks
Machine Learning Researcher / PhD Candidate in Neuroscience
Université de Picardie Jules Verne (UPJV) · Amiens, France
Develop and evaluate machine-learning models using brain-connectivity, neuroimaging and clinical data from approximately 400 stroke patients to investigate cognitive and motor outcomes. Build reproducible Python pipelines covering preprocessing, feature engineering, hyperparameter optimization, cross-validation and interpretation. Apply SVMs, regression, ensemble learning, voxel-level decoding, graph-based methods, statistical and Bayesian analysis. Presented doctoral research at NetSci 2025 in Maastricht and OHBM 2026 in Bordeaux.
Skills
2026 – PresentApplied AI Project
02
Applied AI Project
DocuMind
Lead Developer & AI Architect — Scientific RAG Platform
Applied AI · RAG · Software Engineering
Architected and developed a self-hosted, multi-service platform for organizing scientific PDFs, retrieving evidence and answering grounded questions with citations. Implemented document extraction, semantic chunking, embeddings, vector, keyword and hybrid retrieval with OpenSearch, local model inference, provenance-aware asynchronous processing, authentication, Docker/Nginx deployment and CI/CD.
Skills
View repository2026Scientific Software
03
Scientific Software
PaperMind
Scientific Document Processing & Research Software
Scientific Software · Document Processing
Developed Python and FastAPI services that transform scientific publications into structured, machine-readable packages containing metadata, hierarchical text, references, figures, tables, formulas, provenance and quality information. Designed reusable API and command-line workflows for validation, indexing, retrieval and RAG pipelines.
Skills
View repository2026Robotics Project
04
Robotics Project
Baxter Chess-Playing Robot
Vision-Based Robotic Chess System
Computer Vision · Robotics · Software Engineering
Developed a robotic chess system integrating Kinect-based computer vision, chess-state reasoning, Stockfish, FastAPI, ROS, calibration and motion control. The system detects a human move, updates the board state, generates a legal response and commands a Baxter robot to physically move the appropriate chess piece.
Skills
View repository2025 - 2026Research Project
05
Research Project
GPU-Accelerated Brain Ischemic Stroke Lesion Segmentation
Deep Learning & Medical Imaging
T1-Weighted Brain MRI
Developed deep-learning models for automated ischemic stroke lesion segmentation in T1-weighted brain MRI. Designed and evaluated RAX-NET and VRXU-net, incorporating attention and residual mechanisms. Built GPU-accelerated TensorFlow training pipelines using CUDA and cuDNN for preprocessing, batching, training and experimentation. Conducted model evaluation and optimization on complex stroke lesions; the work resulted in scientific preprints on both architectures.
Skills
View publication2025 - 2026Teaching
06
Teaching
Teaching Python & Machine Learning
Teaching Assistant — Python & Machine Learning
Université de Picardie Jules Verne (UPJV) · Amiens, France
Delivered approximately 60 hours of university-level teaching and laboratory support in Python and machine learning. Supported students with programming, data analysis, model implementation and debugging. Guided hands-on exercises that translated theoretical concepts into working Python implementations, providing individual technical guidance and mentoring during coursework and practical sessions.
Skills
2025Research Collaboration
07
Research Collaboration
Neural Dynamics of Conventional and Virtual Reality Trail Making: A Source-Imaging EEG Study
VR, EEG & Research Software Collaboration
Virtual Reality · EEG · Cognitive Neuroscience
Collaborated on virtual-reality implementations of the Trail Making Test A and B (TMT-A/TMT-B) using a Meta VR headset for cognitive-assessment experiments. Developed and integrated experimental tools for presenting VR-based tasks and supporting EEG signal acquisition. Connected VR interaction, cognitive-task execution and neurophysiological data collection within a research software workflow supporting comparison with conventional Trail Making testing.
Skills
2024Research Project
08
Research Project
Lumbar Spine Degeneration Classification
Deep Learning & Medical Imaging
YOLOv8 · DeepScoreNet
Developed a deep-learning approach for automated classification of lumbar-spine degeneration using YOLOv8 and DeepScoreNet. Built and evaluated a computer-vision pipeline for identifying relevant spinal structures and classifying degenerative conditions. The work resulted in a manuscript released as a medRxiv preprint.
Skills
View publication2023Research Project
09
Research Project
Synthetic Face Generation through Eyes-to-Face Inpainting
Eyes-to-Face Inpainting
Research Project · Published at IEEE IJCB 2023
Developed an eyes-to-face inpainting approach that generates synthetic facial images by combining eye regions from different individuals and reconstructing complete faces. Evaluated synthetic datasets for age and gender estimation and face-recognition training across four benchmark datasets, while investigating demographic biases and representation.
Skills
View publicationJun 2022 – Apr 2023Research Collaboration
10
Research Collaboration
Deep Reinforcement Learning for Secure Wireless Communication
Research Collaborator
Worcester Polytechnic Institute (WPI)
Collaborated on research applying deep reinforcement learning to secure wireless communication. Implemented Deep Deterministic Policy Gradient (DDPG) for continuous decision-making and optimization, and applied the framework to wireless transmission optimization and performance evaluation.
Skills
May 2021 - Oct 2022Industry
11
Industry
Building secure backend services
Backend Developer
Maralhost · Isfahan, Iran
Developed and maintained Django-based backend services and API microservices for web and server-management applications, including relational data models and server-side business logic. Designed authentication and authorization mechanisms including Single Sign-On, developed web-vulnerability assessment tools and supported Linux-based deployment, debugging and collaborative Git workflows.
Skills
2019 - 2023Research Project
12
Research Project
COVID-19 Lung & Lesion Segmentation
Deep Learning, CT Imaging & Web Deployment
Master’s Research Project · Shiraz University of Technology
Developed a deep-learning pipeline for automated lung and COVID-19 lesion segmentation from chest CT scans using clinical imaging data collected from hospitals. Built and trained segmentation models to quantify affected lung regions and integrated preprocessing, inference and visualization into a web-based application designed for practical use by healthcare professionals.
Skills
2019 - 2023AI Project
13
AI Project
Real-Time Face-Mask Detection
Edge AI & Computer Vision
Raspberry Pi Deployment
Developed a real-time deep-learning system for detecting face-mask usage using a Raspberry Pi camera. Implemented image acquisition, preprocessing, inference and visualization for deployment on resource-constrained edge hardware.
Skills
2019 - 2023Education
14
Education
Deep learning meets medical imaging
MSc, Electrical Engineering - Telecommunications
Shiraz University of Technology · Shiraz, Iran
Graduate study focused on machine learning, deep learning and medical image analysis. My thesis investigated quantification of damaged lung tissue due to COVID-19 using computed tomography images and led to practical work in segmentation, computer vision and software deployment.
Skills
2019Research Project
15
Research Project
Persian NLP for E-commerce Review Analysis
Bachelor’s Thesis
University of Isfahan
Developed a Persian-language NLP pipeline using approximately 15,000 customer reviews from Digikala. Built a supervised text-classification model with TF-IDF features and a Linear SVM to classify reviews by purchase recommendation. Implemented sentiment-analysis and text-processing workflows, and applied Latent Dirichlet Allocation (LDA) to identify product-related themes in positive and negative feedback.
Skills
Oct 2017 - Oct 2019Industry
16
Industry
From software to physical control systems
Embedded Systems & Electronics Engineer
Mobarakeh Steel Company · Isfahan, Iran
Analyzed and redesigned an existing electronic control board for rolling-equipment speed regulation, integrating a microcontroller-based control system. Developed control algorithms and evaluated system performance, behavior and reliability under operational conditions.
Skills
2015 - 2019Education
17
Education
A foundation in systems thinking
BEng, Electrical Engineering - Telecommunications
University of Isfahan · Isfahan, Iran
Built a foundation in electrical engineering, mathematics, signals and computation. My final project applied machine learning to Persian-language e-commerce sentiment analysis with positive/negative classification and negative-term extraction.
Skills
ResearchProjectsIndustryEducationTeaching
03 · Expertise
A broad toolkit, organized around outcomes.
My core skills are machine learning, Python, deep learning, computer vision and data science, supported by generative AI, backend engineering and deployment experience.
Machine Learning, Data Science & AI
Statistical analysis and model development for high-dimensional scientific, clinical and real-world data.
- Machine Learning
- Deep Learning
- Computer Vision
- Data Science
- Statistical Modeling
- Bayesian Analysis
- Graph-Based Learning
- Reinforcement Learning
- NLP
- Optimization
Model Development & GPU Computing
Reproducible model development, validation, statistical analysis and visualization in Python.
- Python
- NumPy
- pandas
- Scikit-learn
- XGBoost
- TensorFlow
- Keras
- PyTorch
- CUDA
- cuDNN
- Feature Engineering
- Cross-Validation
- Hyperparameter Optimization
- Model Interpretation
Software Engineering
Backend and full-stack software engineering for APIs, services and maintainable applications.
- Python
- JavaScript
- FastAPI
- Django
- Flask
- Node.js
- Express
- Next.js
- React
- REST APIs
Generative AI & RAG
Grounded AI and document-intelligence systems designed around retrieval quality, context and provenance.
- Generative AI
- LLMs
- RAG
- Embeddings
- Semantic Chunking
- Vector Search
- Keyword Retrieval
- Hybrid Retrieval
- Local Model Inference
- Source-Grounded Question Answering
Backend Infrastructure & MLOps
Data, search and deployment infrastructure for reliable multi-service software systems.
- PostgreSQL
- MongoDB
- Redis
- OpenSearch
- Docker
- Nginx
- Linux
- Git
- GitHub Actions
- CI/CD
Robotics & Experimental Systems
Software integration connecting perception, physical control and cognitive research experiments.
- Robotics
- ROS
- Kinect
- Motion Control
- Robot Calibration
- Virtual Reality
- EEG Data Acquisition
- Embedded Systems
- Raspberry Pi
- System Integration
04 · Research & Publications
Applied research with engineering discipline.
My research spans clinical machine learning, brain connectivity, graph-based methods, medical imaging, computer vision and reproducible scientific computing.
Current doctoral focus
Post-stroke outcomes, brain connectivity & network dynamics
I develop and validate machine-learning models using brain-connectivity and clinical data from approximately 400 stroke patients. My work includes SVMs, regression and ensemble models, voxel- and graph-derived features, Kuramoto simulations, structural-connectivity GNN approaches and reproducible Python pipelines for preprocessing, optimization, cross-validation, statistical analysis, visualization and interpretation.
NetSci 2025 · OHBM 2026Presented doctoral research at NetSci 2025 in Maastricht and the OHBM 2026 Annual Meeting in Bordeaux; manuscript currently under review.
2026arXivVRXU-net: A Deep Learning Approach for Brain Ischemic Stroke Lesion Detection and Segmentation in T1W MRI2025medRxivRAX-NET: Residual Attention Xception Network for Brain Ischemic Stroke Segmentation in T1-Weighted MRI2024medRxivLumbar Spine Degenerative Classification Using YOLO v8 and DeepScoreNet2024arXivEmpowering Medical Imaging with Artificial Intelligence: A Review of Machine Learning Approaches for the Detection, and Segmentation of COVID-19 Using Radiographic and Tomographic Images2023IEEE IJCBSynthetic Face Generation Through Eyes-to-Face Inpainting
05 · Beyond the Code
Communication matters, too.
Teaching
Approximately 60 hours of university-level Python and machine-learning instruction, laboratory support and student guidance.
Presenting
Presented doctoral research at NetSci 2025 in Maastricht and the OHBM 2026 Annual Meeting in Bordeaux to international and multidisciplinary audiences.
Languages
Persian/Farsi (native), English (professional proficiency), French (B1 listening, A2 speaking, A1 writing).
06 · Contact
Let's build something useful.
I'm exploring opportunities in AI Engineering, Machine Learning Engineering, Data Science, Computer Vision and Research Engineering, where analytical thinking and practical system building come together.